Drilling operations system

Abstract

A method can include receiving a query from a drilling operations framework; enhancing the query by retrieving data from a database to form an enhanced query; generating a response to the enhanced query using a large language model; and processing the query, based at least in part on the response, using the drilling operations framework to formulate a control action for drilling operations.

Description

DRILLING OPERATIONS SYSTEMRELATED APPLICATION

[0001] This application claims priority to and the benefit of a U.S. Provisional Application having Serial No. 63 / 562,001 , filed 6 March 2024, which is incorporated by reference herein in its entirety.BACKGROUND

[0002] A resource field may be an accumulation, pool or group of pools of one or more resources (e.g., oil, gas, oil and gas) in a subsurface environment. A resource field may include at least one reservoir. A reservoir may be shaped in a manner that may trap hydrocarbons and may be covered by an impermeable or sealing rock. A bore may be drilled into an environment where the bore may be utilized to form a well that may be utilized in producing hydrocarbons from a reservoir.

[0003] A rig may be a system of components that may be operated to form a bore in an environment, to transport equipment into and out of a bore in an environment, etc. As an example, a rig may include a system that may be used to drill a bore and to acquire information about an environment, about drilling, etc. A resource field may be an onshore field, an offshore field or an on- and offshore field. A rig may include components for performing operations onshore and / or offshore. A rig may be, for example, vessel-based, offshore platform-based, onshore, etc.

[0004] Field planning may occur over one or more phases, which may include an exploration phase that aims to identify and assess an environment (e.g., a prospect, a play, etc.), which may include drilling of one or more bores (e.g., one or more exploratory wells, etc.). Other phases may include appraisal, development and production phases.SUMMARY

[0005] A method can include receiving a query from a drilling operations framework; enhancing the query by retrieving data from a database to form an enhanced query; generating a response to the enhanced query using a large language model; and processing the query, based at least in part on the response, using the drilling operations framework to formulate a control action for drilling operations. Asystem can include a processor; a memory accessible by the processor; processorexecutable instructions stored in the memory and executable to instruct the system to: receive a query from a drilling operations framework; enhance the query by retrieving data from a database to form an enhanced query; generate a response to the enhanced query using a large language model; and process the query, based at least in part on the response, using the drilling operations framework to formulate a control action for drilling operations. One or more computer-readable storage media can include processor-executable instructions to instruct a computing system to: receive a query from a drilling operations framework; enhance the query by retrieving data from a database to form an enhanced query; generate a response to the enhanced query using a large language model; and process the query, based at least in part on the response, using the drilling operations framework to formulate a control action for drilling operations. Various other apparatuses, systems, methods, etc., are also disclosed.

[0006] 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.BRIEF DESCRIPTION OF THE DRAWINGS

[0007] Features and advantages of the described implementations may be more readily understood by reference to the following description taken in conjunction with the accompanying drawings.

[0008] FIG. 1 illustrates examples of equipment in a geologic environment;

[0009] FIG. 2 illustrates examples of equipment and examples of hole types;

[0010] FIG. 3 illustrates an example of a system;

[0011] FIG. 4 illustrates an example of a method and an example of a graphic;

[0012] FIG. 5 illustrates an example of a graphical user interface;

[0013] FIG. 6 illustrates example events;

[0014] FIG. 7 illustrates example events;

[0015] FIG. 8 illustrates example events;

[0016] FIG. 9 illustrates an example of a framework;

[0017] FIG. 10 illustrates an example of a framework;

[0018] FIG. 11 illustrates examples of frameworks;

[0019] FIG. 12 illustrates an example of an architecture of a machine learning model;

[0020] FIG. 13 illustrates an example of an architecture of a retrieval augmented generator;

[0021] FIG. 14 illustrates an example of a system;

[0022] FIG. 15 illustrates an example of a system;

[0023] FIG. 16 illustrates examples of queries and responses;

[0024] FIG. 17 illustrates examples of graphical user interfaces;

[0025] FIG. 18 illustrates an example of a method and an example of a system; and

[0026] FIG. 19 illustrates an example of computing system.DETAILED DESCRIPTION

[0027] The following description includes the best mode presently contemplated for practicing the described implementations. This description is not to be taken in a limiting sense, but rather is made merely for the purpose of describing the general principles of the implementations. The scope of the described implementations should be ascertained with reference to the issued claims.

[0028] FIG. 1 shows an example of a geologic environment 120. In FIG. 1 , the geologic environment 120 may be a sedimentary basin that includes layers (e.g., stratification) that include a reservoir 121 and that may be, for example, intersected by a fault 123 (e.g., or faults). As an example, the geologic environment 120 may be outfitted with a variety of sensors, detectors, actuators, etc. For example, equipment 122 may include communication circuitry to receive and to transmit information with respect to one or more networks 125. Such information may include information associated with downhole equipment 124, which may be equipment to acquire information, to assist with resource recovery, etc. Other equipment 126 may be located remote from a well site and include sensing, detecting, emitting or other circuitry. Such equipment may include storage and communication circuitry to store and to communicate data, instructions, etc. As an example, one or more pieces of equipment may provide for measurement, collection, communication, storage, analysis, etc. of data (e.g., for one or more produced resources, etc.). As an example,one or more satellites may be provided for purposes of communications, data acquisition, etc. For example, FIG. 1 shows a satellite 130 in communication with the network 125 that may be configured for communications, noting that the satellite 130 may additionally or alternatively include circuitry for imagery (e.g., spatial, spectral, temporal, radiometric, etc.).

[0029] FIG. 1 also shows the geologic environment 120 as optionally including equipment 127 and 128 associated with a well that includes a substantially horizontal portion that may intersect with one or more fractures 129. For example, consider a well in a shale formation that may include natural fractures, artificial fractures (e.g., hydraulic fractures) or a combination of natural and artificial fractures. As an example, a well may be drilled for a reservoir that is laterally extensive. In such an example, lateral variations in properties, stresses, etc. may exist where an assessment of such variations may assist with planning, operations, etc. to develop the reservoir (e.g., via fracturing, injecting, extracting, etc.). As an example, the equipment 127 and / or 128 may include components, a system, systems, etc. for fracturing, seismic sensing, analysis of seismic data, assessment of one or more fractures, injection, production, etc. As an example, the equipment 127 and / or 128 may provide for measurement, collection, communication, storage, analysis, etc. of data such as, for example, production data (e.g., for one or more produced resources). As an example, one or more satellites may be provided for purposes of communications, data acquisition, etc.

[0030] FIG. 1 also shows an example of equipment 170 and an example of equipment 180. Such equipment, which may be systems of components, may be suitable for use in the geologic environment 120. While the equipment 170 and 180 are illustrated as land-based, various components may be suitable for use in an offshore system.

[0031] The equipment 170 includes a platform 171 , a derrick 172, a crown block 173, a line 174, a traveling block assembly 175, drawworks 176 and a landing 177 (e.g., a monkeyboard). As an example, the line 174 may be controlled at least in part via the drawworks 176 such that the traveling block assembly 175 travels in a vertical direction with respect to the platform 171. For example, by drawing the line 174 in, the drawworks 176 may cause the line 174 to run through the crown block173 and lift the traveling block assembly 175 skyward away from the platform 171 ; whereas, by allowing the line 174 out, the drawworks 176 may cause the line 174 to run throughthe crown block 173 and lower the traveling block assembly 175 toward the platform 171. Where the traveling block assembly 175 carries pipe (e.g., casing, etc.), tracking of movement of the traveling block 175 may provide an indication as to how much pipe has been deployed.

[0032] A derrick may be a structure used to support a crown block and a traveling block operatively coupled to the crown block at least in part via line. A derrick may be pyramidal in shape and offer a suitable strength-to-weight ratio. A derrick may be movable as a unit or in a piece-by-piece manner (e.g., to be assembled and disassembled).

[0033] As an example, drawworks may include a spool, brakes, a power source and assorted auxiliary devices. Drawworks may controllably reel out and reel in line. Line may be reeled over a crown block and coupled to a traveling block to gain mechanical advantage in a “block and tackle” or “pulley” fashion. Reeling out and in of line may cause a traveling block (e.g., and whatever may be hanging underneath it), to be lowered into or raised out of a bore. Reeling out of line may be powered by gravity and reeling in by a motor, an engine, etc. (e.g., an electric motor, a diesel engine, etc.).

[0034] As an example, a crown block may include a set of pulleys (e.g., sheaves) that may be located at or near a top of a derrick or a mast, over which line is threaded. A traveling block may include a set of sheaves that may be moved up and down in a derrick or a mast via line threaded in the set of sheaves of the traveling block and in the set of sheaves of a crown block. A crown block, a traveling block and a line may form a pulley system of a derrick or a mast, which may enable handling of heavy loads (e.g., drillstring, pipe, casing, liners, etc.) to be lifted out of or lowered into a bore. As an example, line may be about a centimeter to about five centimeters in diameter as, for example, steel cable. Through use of a set of sheaves, such line may carry loads heavier than the line could support as a single strand.

[0035] As an example, a derrickman may be a rig crew member that works on a platform attached to a derrick or a mast. A derrick may include a landing on which a derrickman may stand. As an example, such a landing may be about 10 meters or more above a rig floor. In an operation referred to as trip out of the hole (TOH), a derrickman may wear a safety harness that enables leaning out from the work landing (e.g., monkeyboard) to reach pipe in located at or near the center of a derrick or amast and to throw a line around the pipe and pull it back into its storage location (e.g., fingerboards), for example, until it a time at which it may be desirable to run the pipe back into the bore. As an example, a rig may include automated pipe-handling equipment such that the derrickman controls the machinery rather than physically handling the pipe.

[0036] As an example, a trip may refer to the act of pulling equipment from a bore and / or placing equipment in a bore. As an example, equipment may include a drillstring that may be pulled out of a hole and / or placed or replaced in a hole. As an example, a pipe trip may be performed where a drill bit has dulled or has otherwise ceased to drill efficiently and is to be replaced.

[0037] FIG. 2 shows an example of a wellsite system 200 (e.g., at a wellsite that may be onshore or offshore). As shown, the wellsite system 200 may include a mud tank 201 for holding mud and other material (e.g., where mud may be a drilling fluid), a suction line 203 that serves as an inlet to a mud pump 204 for pumping mud from the mud tank 201 such that mud flows to a vibrating hose 206, a drawworks 207 for winching drill line or drill lines 212, a standpipe 208 that receives mud from the vibrating hose 206, a kelly hose 209 that receives mud from the standpipe 208, a gooseneck or goosenecks 210, a traveling block 211 , a crown block 213 for carrying the traveling block 211 via the drill line or drill lines 212 (see, e.g., the crown block 173 of FIG. 1), a derrick 214 (see, e.g., the derrick 172 of FIG. 1), a kelly 218 or a top drive 240, a kelly drive bushing 219, a rotary table 220, a drill floor 221 , a bell nipple 222, one or more blowout preventers (BOPs) 223, a drillstring 225, a drill bit 226, a casing head 227 and a flow pipe 228 that carries mud and other material to, for example, the mud tank 201.

[0038] In the example system of FIG. 2, a borehole 232 is formed in subsurface formations 230 by rotary drilling; noting that various example embodiments may also use directional drilling.

[0039] As shown in the example of FIG. 2, the drillstring 225 is suspended within the borehole 232 and has a drillstring assembly 250 that includes the drill bit 226 at its lower end. As an example, the drillstring assembly 250 may be a bottom hole assembly (BHA).

[0040] The wellsite system 200 may provide for operation of the drillstring 225 and other operations. As shown, the wellsite system 200 includes the platform 211and the derrick 214 positioned over the borehole 232. As mentioned, the wellsite system 200 may include the rotary table 220 where the drillstring 225 pass through an opening in the rotary table 220.

[0041] As shown in the example of FIG. 2, the wellsite system 200 may include the kelly 218 and associated components, etc., or a top drive 240 and associated components. As to a kelly example, the kelly 218 may be a square or hexagonal metal / alloy bar with a hole drilled therein that serves as a mud flow path. The kelly 218 may be used to transmit rotary motion from the rotary table 220 via the kelly drive bushing 219 to the drillstring 225, while allowing the drillstring 225 to be lowered or raised during rotation. The kelly 218 may pass through the kelly drive bushing 219, which may be driven by the rotary table 220. As an example, the rotary table 220 may include a master bushing that operatively couples to the kelly drive bushing 219 such that rotation of the rotary table 220 may turn the kelly drive bushing 219 and hence the kelly 218. The kelly drive bushing 219 may include an inside profile matching an outside profile (e.g., square, hexagonal, etc.) of the kelly 218; however, with slightly larger dimensions so that the kelly 218 may freely move up and down inside the kelly drive bushing 219.

[0042] As to a top drive example, the top drive 240 may provide functions performed by a kelly and a rotary table. The top drive 240 may turn the drillstring 225. As an example, the top drive 240 may include one or more motors (e.g., electric and / or hydraulic) connected with appropriate gearing to a short section of pipe called a quill, that in turn may be screwed into a saver sub or the drillstring 225 itself. The top drive 240 may be suspended from the traveling block 211 , so the rotary mechanism is free to travel up and down the derrick 214. As an example, a top drive 240 may allow for drilling to be performed with more joint stands than a kelly / rotary table approach.

[0043] In the example of FIG. 2, the mud tank 201 may hold mud, which may be one or more types of drilling fluids. As an example, a wellbore may be drilled to produce fluid, inject fluid or both (e.g., hydrocarbons, minerals, water, etc.).

[0044] In the example of FIG. 2, the drillstring 225 (e.g., including one or more downhole tools) may be composed of a series of pipes threadably connected together to form a long tube with the drill bit 226 at the lower end thereof. As the drillstring 225 is advanced into a wellbore for drilling, at some point in time prior to or coincident with drilling, the mud may be pumped by the pump 204 from the mud tank 201 (e.g., orother source) via the lines 206, 208 and 209 to a port of the kelly 218 or, for example, to a port of the top drive 240. The mud may then flow via a passage (e.g., or passages) in the drillstring 225 and out of ports located on the drill bit 226 (see, e.g., a directional arrow). As the mud exits the drillstring 225 via ports in the drill bit 226, it may then circulate upwardly through an annular region between an outer surface(s) of the drillstring 225 and surrounding wall(s) (e.g., open borehole, casing, etc.), as indicated by directional arrows. In such a manner, the mud lubricates the drill bit 226 and carries heat energy (e.g., frictional or other energy) and formation cuttings to the surface where the mud (e.g., and cuttings) may be returned to the mud tank 201 , for example, for recirculation (e.g., with processing to remove cuttings, etc.).

[0045] The mud pumped by the pump 204 into the drillstring 225 may, after exiting the drillstring 225, form a mudcake that lines the wellbore which, among other functions, may reduce friction between the drillstring 225 and surrounding wall(s) (e.g., borehole, casing, etc.). A reduction in friction may facilitate advancing or retracting the drillstring 225. During a drilling operation, the entire drillstring 225 may be pulled from a wellbore and optionally replaced, for example, with a new or sharpened drill bit, a smaller diameter drillstring, etc. As mentioned, the act of pulling a drillstring out of a hole or replacing it in a hole is referred to as tripping. A trip may be referred to as an upward trip or an outward trip or as a downward trip or an inward trip depending on trip direction.

[0046] As an example, consider a downward trip where upon arrival of the drill bit 226 of the drillstring 225 at a bottom of a wellbore, pumping of the mud commences to lubricate the drill bit 226 for purposes of drilling to enlarge the wellbore. As mentioned, the mud may be pumped by the pump 204 into a passage of the drillstring 225 and, upon filling of the passage, the mud may be used as a transmission medium to transmit energy, for example, energy that may encode information as in mud-pulse telemetry.

[0047] As an example, mud-pulse telemetry equipment may include a downhole device configured to effect changes in pressure in the mud to create an acoustic wave or waves upon which information may modulated. In such an example, information from downhole equipment (e.g., one or more modules of the drillstring 225) may be transmitted uphole to an uphole device, which may relay such information to other equipment for processing, control, etc.

[0048] As an example, telemetry equipment may operate via transmission of energy via the drillstring 225 itself. For example, consider a signal generator that imparts coded energy signals to the drillstring 225 and repeaters that may receive such energy and repeat it to further transmit the coded energy signals (e.g., information, etc.).

[0049] As an example, the drillstring 225 may be fitted with telemetry equipment 252 that includes a rotatable drive shaft, a turbine impeller mechanically coupled to the drive shaft such that the mud may cause the turbine impeller to rotate, a modulator rotor mechanically coupled to the drive shaft such that rotation of the turbine impeller causes said modulator rotor to rotate, a modulator stator mounted adjacent to or proximate to the modulator rotor such that rotation of the modulator rotor relative to the modulator stator creates pressure pulses in the mud, and a controllable brake for selectively braking rotation of the modulator rotor to modulate pressure pulses. In such example, an alternator may be coupled to the aforementioned drive shaft where the alternator includes at least one stator winding electrically coupled to a control circuit to selectively short the at least one stator winding to electromagnetically brake the alternator and thereby selectively brake rotation of the modulator rotor to modulate the pressure pulses in the mud.

[0050] In the example of FIG. 2, an uphole control and / or data acquisition system 262 may include circuitry to sense pressure pulses generated by telemetry equipment 252 and, for example, communicate sensed pressure pulses or information derived therefrom for process, control, etc.

[0051] The assembly 250 of the illustrated example includes a logging-while- drilling (LWD) module 254 (e.g., a LWD tool), a measuring-while-drilling (MWD) module 256 (e.g., a MWD tool), an optional module 258, a roto-steerable system (RSS) and / or motor 260, and the drill bit 226. Such components or modules may be referred to as tools where a drillstring may include a plurality of tools.

[0052] As to an RSS, it involves technology utilized for directional drilling. Directional drilling involves drilling into the Earth to form a deviated bore such that the trajectory of the bore is not vertical; rather, the trajectory deviates from vertical along one or more portions of the bore. As an example, consider a target that is located at a lateral distance from a surface location where a rig may be stationed. In such an example, drilling may commence with a vertical portion and then deviate from verticalsuch that the bore is aimed at the target and, eventually, reaches the target. Directional drilling may be implemented where a target may be inaccessible from a vertical location at the surface of the Earth, where material exists in the Earth that may impede drilling or otherwise be detrimental (e.g., consider a salt dome, etc.), where a formation is laterally extensive (e.g., consider a relatively thin yet laterally extensive reservoir), where multiple bores are to be drilled from a single surface bore, where a relief well is desired, etc.

[0053] One approach to directional drilling involves a mud motor; however, a mud motor may present some challenges depending on factors such as rate of penetration (ROP), transferring weight to a bit (e.g., weight on bit, WOB) due to friction, etc. A mud motor may be a positive displacement motor (PDM) that operates to drive a bit (e.g., during directional drilling, etc.). A PDM operates as drilling fluid is pumped through it where the PDM converts hydraulic power of the drilling fluid into mechanical power to cause the bit to rotate.

[0054] As an example, a PDM may operate in a combined rotating mode where surface equipment is utilized to rotate a bit of a drillstring (e.g., a rotary table, a top drive, etc.) by rotating the entire drillstring and where drilling fluid is utilized to rotate the bit of the drillstring. In such an example, a surface RPM (SRPM) may be determined by use of the surface equipment and a downhole RPM of the mud motor may be determined using various factors related to flow of drilling fluid, mud motor type, etc. As an example, in the combined rotating mode, bit RPM may be determined or estimated as a sum of the SRPM and the mud motor RPM, assuming the SRPM and the mud motor RPM are in the same direction.

[0055] As an example, a PDM mud motor may operate in a so-called sliding mode, when the drillstring is not rotated from the surface to drive a drill bit in a particular cutting direction. In such an example, a bit RPM may be determined or estimated based on the RPM of the mud motor. As an example, during a sliding mode, oscillation of a drillstring may be provided by surface equipment, for example, to oscillate the drillstring in a clockwise and a counter-clockwise direction, which may, for example, help to reduce risk of sticking, etc.

[0056] An RSS may drill directionally where there is continuous rotation from surface equipment, which may alleviate the sliding of a steerable motor (e.g., a PDM). An RSS may be deployed when drilling directionally (e.g., deviated, horizontal, orextended-reach wells). An RSS may aim to minimize interaction with a borehole wall, which may help to preserve borehole quality. An RSS may aim to exert a relatively consistent side force akin to stabilizers that rotate with the drillstring or orient the bit in the desired direction while continuously rotating at the same number of rotations per minute as the drillstring.

[0057] The LWD module 254 may be housed in a suitable type of drill collar and may contain one or a plurality of selected types of logging tools. It will also be understood that more than one LWD and / or MWD module may be employed. Where the position of a module is mentioned, as an example, it may refer to a module at the position of the LWD module 254, the MWD module 256, etc. An LWD module may include capabilities for measuring, processing, and storing information, as well as for communicating with the surface equipment. In the illustrated example, the LWD module 254 may include a seismic measuring device.

[0058] The MWD module 256 may be housed in a suitable type of drill collar and may contain one or more devices for measuring characteristics of the drillstring 225 and the drill bit 226. As an example, the MWD module 256 may include equipment for generating electrical power, for example, to power various components of the drillstring 225. As an example, the MWD module 256 may include the telemetry equipment 252, for example, where the turbine impeller may generate power by flow of the mud; it being understood that other power and / or battery systems may be employed for purposes of powering various components. As an example, the MWD module 256 may include one or more of the following types of measuring devices: a weight-on-bit measuring device, a torque measuring device, a vibration measuring device, a shock measuring device, a stick slip measuring device, a direction measuring device, and an inclination measuring device.

[0059] FIG. 2 also shows some examples of types of holes that may be drilled. For example, consider a slant hole 272, an S-shaped hole 274, a deep inclined hole 276 and a horizontal hole 278.

[0060] As an example, a drilling operation may include directional drilling where, for example, at least a portion of a well includes a curved axis. For example, consider a radius that defines curvature where an inclination with regard to the vertical may vary until reaching an angle between about 30 degrees and about 60 degrees or, for example, an angle to about 90 degrees or possibly greater than about 90 degrees.

[0061] As an example, a directional well may include several shapes where each of the shapes may aim to meet particular operational demands. As an example, a drilling process may be performed on the basis of information as and when it is relayed to a drilling engineer. As an example, inclination and / or direction may be modified based on information received during a drilling process.

[0062] As an example, deviation of a bore may be accomplished in part by use of one or more of an RSS, a downhole motor and / or a turbine. As to a motor, for example, a drillstring may include a positive displacement motor (PDM).

[0063] As an example, a system may be a steerable system and include equipment to perform a method such as geosteering. As an example, a steerable system may include a PDM or a turbine on a lower part of a drillstring which, just above a drill bit, a bent sub may be mounted. As an example, above a PDM, MWD equipment that provides real time or near real time data of interest (e.g., inclination, direction, pressure, temperature, real weight on the drill bit, torque stress, etc.) and / or LWD equipment may be installed. As to the latter, LWD equipment may make it possible to send to the surface various types of data of interest, including for example, geological data (e.g., gamma ray log, resistivity, density and sonic logs, etc.).

[0064] The coupling of sensors providing information on the course of a well trajectory, in real time or near real time, with, for example, one or more logs characterizing the formations from a geological viewpoint, may allow for implementing a geosteering method. Such a method may include navigating a subsurface environment, for example, to follow a desired route to reach a desired target or targets.

[0065] As an example, a drillstring may include an azimuthal density neutron (ADN) tool for measuring density and porosity; a MWD tool for measuring inclination, azimuth and shocks; a compensated dual resistivity (CDR) tool for measuring resistivity and gamma ray related phenomena; one or more variable gauge stabilizers; one or more bend joints; and a geosteering tool, which may include a motor and optionally equipment for measuring and / or responding to one or more of inclination, resistivity and gamma ray related phenomena.

[0066] As an example, geosteering may include intentional directional control of a wellbore based on results of downhole geological logging measurements in a manner that aims to keep a directional wellbore within a desired region, zone (e.g., a pay zone), etc. As an example, geosteering may include directing a wellbore to keepthe wellbore in a particular section of a reservoir, for example, to minimize gas and / or water breakthrough and, for example, to maximize economic production from a well that includes the wellbore.

[0067] Referring again to FIG. 2, the wellsite system 200 may include one or more sensors 264 that are operatively coupled to the control and / or data acquisition system 262. As an example, a sensor or sensors may be at surface locations. As an example, a sensor or sensors may be at downhole locations. As an example, a sensor or sensors may be at one or more remote locations that are not within a distance of the order of about one hundred meters from the wellsite system 200. As an example, a sensor or sensor may be at an offset wellsite where the wellsite system 200 and the offset wellsite are in a common field (e.g., oil and / or gas field).

[0068] As an example, one or more of the sensors 264 may be provided for tracking pipe, tracking movement of at least a portion of a drillstring, etc.

[0069] As an example, the system 200 may include one or more sensors 266 that may sense and / or transmit signals to a fluid conduit such as a drilling fluid conduit (e.g., a drilling mud conduit). For example, in the system 200, the one or more sensors 266 may be operatively coupled to portions of the standpipe 208 through which mud flows. As an example, a downhole tool may generate pulses that may travel through the mud and be sensed by one or more of the one or more sensors 266. In such an example, the downhole tool may include associated circuitry such as, for example, encoding circuitry that may encode signals, for example, to reduce demands as to transmission. As an example, circuitry at the surface may include decoding circuitry to decode encoded information transmitted at least in part via mud-pulse telemetry. As an example, circuitry at the surface may include encoder circuitry and / or decoder circuitry and circuitry downhole may include encoder circuitry and / or decoder circuitry. As an example, the system 200 may include a transmitter that may generate signals that may be transmitted downhole via mud (e.g., drilling fluid) as a transmission medium.

[0070] As an example, one or more portions of a drillstring may become stuck. The term stuck may refer to one or more of varying degrees of inability to move or remove a drillstring from a bore. As an example, in a stuck condition, it might be possible to rotate pipe or lower it back into a bore or, for example, in a stuck condition, there may be an inability to move the drillstring axially in the bore, though some amountof rotation may be possible. As an example, in a stuck condition, there may be an inability to move at least a portion of the drillstring axially and rotationally.

[0071] As to the term “stuck pipe”, this term may refer to a portion of a drillstring that cannot be rotated or moved axially. As an example, a condition referred to as “differential sticking” may be a condition whereby the drillstring cannot be moved (e.g., rotated or reciprocated) along the axis of the bore. Differential sticking may occur when high-contact forces caused by low reservoir pressures, high wellbore pressures, or both, are exerted over a sufficiently large area of the drillstring. Differential sticking may have time and financial cost.

[0072] As an example, a sticking force may be a product of the differential pressure between the wellbore and the reservoir and the area that the differential pressure is acting upon. This means that a relatively low differential pressure (delta p) applied over a large working area may be just as effective in sticking pipe as may a high differential pressure applied over a small area.

[0073] As an example, a condition referred to as “mechanical sticking” may be a condition where limiting or prevention of motion of the drillstring by a mechanism other than differential pressure sticking occurs. Mechanical sticking may be caused, for example, by one or more of junk in the hole, wellbore geometry anomalies, cement, keyseats or a buildup of cuttings in the annulus.

[0074] Various types of data associated with field operations may be 1 -D series data. For example, consider data as to one or more of a drilling system, downhole states, formation attributes, and surface mechanics being measured as single or multichannel time series data.

[0075] FIG. 3 shows an example of various components of a hoisting system 300, which includes a cable 301 , a drawworks 310, a traveling block 311 , a hook 312, a crown block 313, a top drive 314, a drillstring 316, a cable deadline tiedown anchor 320, a cable supply reel 330, one or more sensors 340 and circuitry 350 operatively coupled to the one or more sensors 340. In the example of FIG. 3, the hoisting system 300 may include various sensors, which may include one or more of load sensors, displacement sensors, accelerometers, etc. As an example, the cable deadline tiedown anchor 320 may be fit with a load cell (e.g., a load sensor).

[0076] The hoisting system 300 may be part of a wellsite system (see, e.g., FIG.1 and FIG. 2). In such a system, a measurement channel may be a block positionmeasurement channel, referred to as BPOS, which provides measurements of a height of a traveling block, which may be defined about a deadpoint (e.g., zero point) and may have deviations from that deadpoint in positive and / or negative directions. For example, consider a traveling block that may move in a range of approximately -5 meters to +45 meters, for a total excursion of approximately 50 meters. As an example, a null point or deadpoint may be defined to make a scale positive, negative or both positive and negative. In such an example, a rig height may be greater than approximately 50 meters (e.g., a crown block may be set at a height from the ground or rig floor in excess of approximately 50 meters). While various examples are given for land-based field operations (e.g., fixed, truck-based, etc.), various methods may apply for marine-based operations (e.g., vessel-based rigs, platform rigs, etc.).

[0077] As to the distance range of a traveling block, it may be sufficient for adding and removing drill pipe and / or other components. As an example, a stand may be two or three single joints of drill pipe or drill collars that remain screwed together during tripping operations. As an example, a stand may be a three-joint stand. As an example, a drill pipe length may be approximately 10 m (e.g., consider from about 27 ft to about 30 ft); noting that shorter or longer drill pipe may be utilized. Where a stand is composed of three lengths of approximately 10 m drill pipe, the stand may have an over length of approximately 30 m (e.g., approximately 100 ft). As such, a traveling block that has a total excursion of approximately 50 m may be raised and lowered to accommodate a stand of approximately 30 m (e.g., for addition to a drillstring or removal from a drillstring).

[0078] BPOS is a type of real-time channel that reflects surface mechanical properties of a rig. Another example of a channel is hook load, which may be referred to as HKLD. HKLD may be a 1-D series measurement of the load of a hook. As to a derivative, a first derivative may be a load velocity and a second derivative may be a load acceleration. Such data channels may be utilized to infer and monitor various operations and / or conditions. In some examples, a rig may be represented as being in one or more states, which may be referred to as rig states.

[0079] As to the HKLD channel, it may help to detect if a rig is “in slips”, while the BPOS channel may be a primary channel for depth tracking during drilling. For example, BPOS may be utilized to determine a measured depth in a geologicenvironment (e.g., a borehole being drilled, etc.). As to the condition or state “in slips”, HKLD is at a much lower value than in the condition or state “out of slips”.

[0080] The term slips refers to a device or assembly that may be used to grip a drillstring (e.g., drill collar, drill pipe, etc.) in a relatively non-damaging manner and suspend it in a rotary table. Slips may include three or more steel wedges that are hinged together, forming a near circle around a drill pipe. On the drill pipe side (inside surface), the slips are fitted with replaceable, hardened tool steel teeth that embed slightly into the side of the pipe. The outsides of the slips are tapered to match the taper of the rotary table. After the rig crew places the slips around the drill pipe and in the rotary, a driller may control a rig to slowly lower the drillstring. As the teeth on the inside of the slips grip the pipe, the slips are pulled down. This downward force pulls the outer wedges down, providing a compressive force inward on the drill pipe and effectively locking components together. Then the rig crew may unscrew the upper portion of the drillstring (e.g., a kelly, saver sub, a joint or stand of pipe) while the lower part is suspended. After some other component is screwed onto the lower part of the drillstring, the driller raises the drillstring to unlock the gripping action of the slips, and a rig crew may remove the slips from the rotary.

[0081] A hook load sensor may be used to measure a weight of load on a drillstring and may be used to detect whether a drillstring is in-slips or out-of-slips. When the drillstring is in-slips, motion from the blocks or motion compensator do not have an effect on the depth of a drill bit at the end of the drillstring (e.g., it will tend to remain stationary). Where movement of a traveling block is via a drawworks encoder (DWE), which may be mounted on a shaft of the drawworks, acquired DWE information (e.g., BPOS) does not augment the recorded drill bit depth. When a drillstring is out-of-slips (e.g., drilling ahead), DWE information (e.g., BPOS) may augment the recorded bit depth. The difference in hook load weight (HKLD) between in-slips and out-of-slips tends to be distinguishable. As to marine operations, heave of a vessel may affect bit depth whether a drillstring is in-slips or out-of-slips. As an exmaple, a vessel may include one or more heave sensors, which may sense data that may be recorded as 1-D series data.

[0082] As to marine operations, a vessel may expeirence various types of motion, such as, for example, one or more of heave, sway and surge. Heave is a linear vertical (up / down) motion, sway is linear lateral (side-to-side or port-starboard)motion, and surge is linear longitudinal (front / back or bow / stern) motion imparted by maritime conditions. As an exmaple, a vessel may include one or more heave sensors, one or more sway sensors and / or one or more surge sensors, each of which may sense data that may be recorded as 1 -D series data.

[0083] As an exmaple, BPOS alone, or combined with one or more other channels, may be used to detect whether a rig is “on bottom” drilling or “tripping”, etc. An inferred state may be further consumed by one or more systems such as, for example, an automatic drilling control system, which may be a dynamic field operations system or a part thereof. In such an exmaple, the conditions, operations, states, etc., as discerned from BPOS and / or other channel data may be predicates to making one or more drilling decisions, which may include one or more control decisions (e.g., of a controller that is operatively coupled to one or more pieces of field equipment, etc.).

[0084] A block may be a set of pulleys used to gain mechanical advantage in lifting or dragging heavy objects. There may be two blocks on a drilling rig, the crown block and the traveling block. Each may include several sheaves that are rigged with steel drilling cable or line such that the traveling block may be raised (or lowered) by reeling in (or out) a spool of drilling line on the drawworks. As such, block position may refer to the position of the traveling block, which may vary with respect to time. FIG. 1 shows the traveling block assembly 175, FIG. 2 shows the traveling block 211 and FIG. 3 shows the traveling block 311 .

[0085] A hook may be high-capacity J-shaped equipment used to hang various equipment such as a swivel and kelly, elevator bails, or a topdrive. FIG. 3 shows the hook 312 as operatively coupled to a top drive 314. As shown in FIG. 2, a hook may be attached to the bottom of the traveling block 211 (e.g., part of the traveling block assembly 175 of FIG. 1). A hook may provide a way to pick up heavy loads with a traveling block. The hook may be either locked (e.g., a normal condition) or free to rotate, so that it may be mated or decoupled with items positioned around the rig floor, etc.

[0086] Hook load may be the total force pulling down on a hook as carried by a traveling block. The total force includes the weight of the drillstring in air, the drill collars and ancillary equipment, reduced by forces that tend to reduce that weight. Some forces that might reduce the weight include friction along a bore wall (especiallyin deviated wells) and buoyant forces on a drillstring caused by its immersion in drilling fluid (e.g., and / or other fluid). If a blowout preventer (BOP) (e.g., or BOPs) is closed, pressure in a bore acting on cross-sectional area of a drillstring in the BOP may also exert an upward force.

[0087] A standpipe may be a rigid metal conduit that provides a high-pressure pathway for drilling fluid to travel approximately one-third of the way up the derrick, where it connects to a flexible high-pressure hose (e.g., kelly hose). A large rig may be fitted with more than one standpipe so that downtime is kept to a minimum if one standpipe demands repair. FIG. 2 shows the standpipe 208 as being a conduit for drilling fluid (e.g., drilling mud, etc.). Pressure of fluid within the standpipe 208 may be referred to as standpipe pressure.

[0088] As to surface torque, such a measurement may be provided by equipment at a rig site. As an example, one or more sensors may be utilized to measure surface torque, which may provide for direct and / or indirect measurement of surface torque associated with a drillstring. As an example, equipment may include a drill pipe torque measurement and controller system with one or more of analog frequency output and digital output. As an example, a torque sensor may be associated with a coupling that includes a resilient element operatively joining an input element and an output element where the resilient element allows the input and output elements to twist with respect to one another in response to torque being transmitted through the torque sensor where the twisting may be measured and used to determine the torque being transmitted. As an example, such a coupling may be located between a drive and drill pipe. As an example, torque may be determined via an inertia sensor or sensors. As an example, equipment at a rig site may include one or more sensors for measurement and / or determination of torque (e.g., in units of Nm, etc.).

[0089] As an example, equipment may include a real-time drilling service system that may provide data such as weight transfer information, torque transfer information, equivalent circulation density (ECD) information, downhole mechanical specific energy (DMSE) information, motion information (e.g., as to stall, stick-slip, etc.), bending information, vibrational amplitude information (e.g., axial, lateral and / or torsional), rate of penetration (ROP) information, pressure information, differential pressure information, flow information, etc. As an example, sensor information may include inclination, azimuth, total vertical depth, etc. As an example, a system mayprovide information as to whirl (e.g., backward whirl, etc.) and may optionally provide information such as one or more alerts (e.g., “severe backward whirl: stop and restart with lower surface RPM”, etc.).

[0090] As to DMSE, it may be a MSE as associated with downhole energy. MSE may be utilized as a measure of drilling efficiency. MSE may be defined as the energy required to remove a unit volume of rock. For optimal drilling efficiency, field operations may aim to minimize the MSE and to maximize ROP. As an example, to control MSE, field equipment may be controlled as to factors such as, for example, one or more of WOB, torque, ROP and RPM.

[0091] A drill bit may be defined as a tool used to crush and / or cut rock. As explained, various rig equipment may directly and / or indirectly assist a drill bit in crushing and / or cutting the rock. Various drill bits may work by scraping or crushing the rock, or both, usually as part of a rotational motion; noting that some bits, known as hammer bits, pound rock. During drilling, various equipment may be controlled to deliver energy to a drill bit to crush and / or cut rock to thereby lengthen a borehole. As explained, drilling may aim to minimize MSE and maximize ROP while maintaining borehole quality (e.g., integrity, etc.). As an example, various equipment may be controlled as to energy delivered to a drillstring and / or a drill bit, for example, to address one or more conditions, which may include, for example, one or more conditions that may cause sticking of a drillstring and / or increase risk of sticking of a drillstring and / or one or more conditions involving actual sticking of a drillstring (e.g., getting a drillstring unstuck, etc.). As various physical interactions may occur between a drillstring and a formation (e.g., a borehole wall), controlled delivery of energy, material(s) (e.g., drilling fluid additives, etc.), etc., may provide for reduced risk of damage to the drillstring and / or the formation.

[0092] As explained, a drillstring may include a tool or tools that include various sensors that may make various measurements. For example, consider the OPTIDRILL tool (SLB, Houston, Texas), which includes strain gauges, accelerometers, magnetometer(s), gyroscope(s), etc. For example, such a tool may acquire weight on bit measurements (WOB) using a strain gauge (e.g., 10 second moving window with bandwidth of 200 Hz), torque measurements using a strain gauge (e.g., 10 second moving window with bandwidth of 200 Hz), bending moment using a strain gauge (e.g., 10 second moving window with bandwidth of 200 Hz), vibrationusing one or more accelerometers (e.g., 30 second RMS with bandwidth of 0.2 to 150 Hz), rotational speed using a magnetometer and a gyroscope (e.g., 30 moving window with bandwidth of 4 Hz), annular and internal pressures using one or more strain gauges (e.g., 1 second average with bandwidth of 200 Hz), annular and internal temperatures using one or more temperature sensors (1 second average with bandwidth of 10 Hz), and continuous inclination using an accelerometer (30 second average with bandwidth of 10 Hz).

[0093] As mentioned, channels of real time drilling operation data may be received and characterized using generated synthetic data, which may be generated based at least in part on one or more operational parameters associated with the real time drilling operation. Such real time drilling operation data may include surface data and / or downhole data. As mentioned, data availability may differ temporally (e.g., frequency, gaps, etc.) and / or otherwise (e.g., resolution, etc.). Such data may differ as to noise level and / or noise characteristics. While various types of sensors are mentioned, equipment may be utilized that may not include one or more types of downhole sensors. In such instances, a method may be utilized that may determine one or more downhole values.

[0094] FIG. 4 shows an example of a method 400 that includes various blocks that may receive data, perform one or more analyses, perform one or more decisions, etc., to determine one or more states. In the example of FIG. 4, various examples of states may be illustrated with respect to shading, color, etc., for example, shading of various blocks may be utilized as a key for a graphical display (e.g., a graphical user interface), as shown in FIG. 4. In FIG. 4, the example states include drilling, nondrilling, run-in-hole (RIH), pull-out-of-hole (POOH), pre-connection, connection, postconnection, and absent.

[0095] In FIG. 4, drilling is drilling to increase the length of a wellbore. Nondrilling activity may be determined to be occurring when no other activities are occurring (e.g., drilling, RIH, POOH, pre-connection, connection, post-connection) and where the end of a current drill stand has not yet been reached. During non-drilling, the flow rate of fluid (e.g., mud) being pumped into a drillstring may increase and / or decrease, the rate of rotation of a drillstring may increase and / or decrease, a downhole tool (e.g, a drill bit) may move upwards and / or downwards, or a combination thereof.A non-drilling activity may be or include a time when a drill bit is idle (e.g., not drilling) and a slips assembly is not engaged with a drillstring.

[0096] Pre-connection may be where a downhole tool (e.g., a drill bit) has completed drilling operations for a current section of pipe, but the slips assembly has not begun to move (e.g., radially-inward) into engagement with the drillstring. During pre- connection, the flow rate of fluid being pumped into the drillstring may increase and / or decrease, the rate of rotation of the drillstring may increase and / or decrease, the downhole tool (e.g., the drill bit) may move upwards and / or downwards, or a combination thereof.

[0097] Connection may be where a slips assembly is engaged with, and supports, a drillstring (e.g., the drillstring is “in-slips”). When a connection is occurring, a segment (e.g., a pipe, a stand, etc.) may be added to the drillstring to increase the length of the drillstring, or a segment may be removed from the drillstring to reduce the length of the drillstring.

[0098] Post-connection may be where the drillstring is released by a slips assembly, and a downhole tool (e.g., the drill bit) are lowered to be on-bottom (e.g., bottom of hole or BOH). During post-connection, the flow rate of fluid being pumped into a drillstring may increase and / or decrease, the rate of rotation of a drillstring may increase and / or decrease, a downhole tool (e.g., the drill bit) may move upwards and / or downwards, or a combination thereof.

[0099] As to an absent state, it may indicate a scenario where data are not being received (e.g., at least one of a plurality of inputs is missing).

[0100] As an example, a method may be utilized to determine a slips status. For example, slips status may include one or more of the following: In-slips where a slips assembly is engaged with, and supports, a drillstring (“in-slips”); out-of-slips where the slips assembly is not engaged with, and does not support, the drillstring; and absent where data are not received (e.g., at least one of the inputs is missing).

[0101] The method 400 of FIG. 4 may include various data acquisition or data reception blocks 402, 406, 408, etc., various decision block 405, 407, 409, 413, 415, 417, and 443, detection blocks 412 and 442 and state blocks. As an example, a block or blocks may provide for processing data, which may include real-time data. For example, a block 404 may provide for identifying one or more gaps and filling one or more of the one or more gaps (e.g., via interpolation, via insertion of data valuesindicative of missing values, etc.). As to the decision block 409, it may decide whether a drilling section is detected 412 or a non-drilling section is detected 442. For example, a drilling section may be indicative of a drilling state, a non-drilling state, a postconnection state, a pre-connection state, etc.; whereas, a non-drilling section may be indicative of a tripping operation such as, for example, RIH or POOH. As to the decision block 407, it may provide for detection of a connection state. As shown, various decision blocks may be implemented to detect a state.

[0102] As an example, in the method 400, measurements (e.g., data) may include a depth of a wellbore (e.g., a measured depth), a depth of a drill bit (e.g., a measured depth), a position of a travelling block (e.g., BPOS), or a combination thereof. A set of measurements may or may not include weight on hook (e.g., HKLD), or weight on a drill bit (e.g., WOB). Each set of measurements may be captured / received a predetermined amount of time after a previous set of measurements is captured / received. A predetermined amount of time may be, for example, about three seconds; however, the predetermined amount of time may be shorter or longer.

[0103] FIG. 5 shows an example GUI 500 where comparisons may be made for pickup (PU) and slackoff (SO) weights taken during connections using a broomstick model. The example GUI 500 may render broomstick model plots with respect to depth (e.g., measured depth, etc.). Where a borehole is vertical, plotting with respect to depth may provide for some insight as the direction of the acceleration of gravity is vertical. Thus, an operator may understand how gravity impacts friction with respect to a drillstring, a BHA, a bit, drilling fluid (e.g., mud), etc. Further, pickup (PU) and slackoff (SO) may be with respect to gravity downhole, not just at surface.

[0104] As an example, a GUI may provide for rendering one or more broomstick model plots with respect to time (e.g., horizontally, vertically, etc.). In such an example, a broomstick model plot may be utilized to ascertain one or more friction factors with respect to time. As an example, a broomstick plot or broomstick model plot (e.g., a plot of model results, etc.), may be a full broomstick plot, a half broomstick plot or another portion of a broomstick plot. For example, where PU and SO are concerned, they may correspond to different directions such that a full broomstick plot may be generated; noting that a half broomstick plot for PU and / or a half broomstick plot for SO may be generated. As to TQLS, where the torque is in a particularrotational direction (e.g., a rotational direction of a bit for drilling), a broomstick plot may be a half broomstick plot; noting that torque may be acquired in two rotational directions (e.g., clockwise and counterclockwise), which may provide for rendering a plot in a full broomstick manner.

[0105] As an example, a system may provide for real-time (RT) torque and drag(T&D) monitoring. Abnormal torque and drag, which commonly refers to overpull, underpull, and high-torque load, are indications of excess frictional effects between the drillstring and the wellbore. Various conditions may cause these effects, including tight hole, differential sticking, poor hole cleaning, keyseats, etc. Failing to detect these anomalies may cause excessive wear on a drillstring and may eventually lead to severe stuck pipe conditions.

[0106] FIG. 6 shows some example events 600 that include differential sticking, geopressure, unconsolidated zone, fracture or faulted zone, undergauge hole, and seating. As shown in FIG. 6, the example events 600 can be associated with interactions between a drillstring and a formation, drillstring related operations and a drillstring, drillstring related operations and a formation, etc.

[0107] FIG. 7 shows some example events 700 that include reactive formation, mobile formation, collapsed casing, junk in hole, cement-related and drillstring vibration. As shown in FIG. 7, the example events 700 can be associated with interactions between a drillstring and a formation, drillstring related operations and a drillstring, drillstring related operations and a formation, etc.

[0108] FIG. 8 shows some example events 800 that include a wellbore geometry event and a poor hole cleaning event. As shown in FIG. 8, the example events 800 can be associated with interactions between a drillstring and a formation, drillstring related operations and a drillstring, drillstring related operations and a formation, etc.

[0109] As an example, loose or unconsolidated formation sands or gravels can collapse into a borehole and pack-off a drillstring as supporting rock is removed by a bit. Schists, laminated shales, fractures and faults can create loose rock that caves into the hole and jam a drillstring. As an example, one or more of such factors may be taken into consideration by a framework such as the framework 600 (e.g., as inputs, models, etc ).

[0110] In regions where tectonic stresses are high, rock is being deformed by movement of the Earth's crust. In such areas, the rock around a bore may collapse into the bore. In some cases, hydrostatic pressure to stabilize a hole may be much higher than the fracture initiation pressure of exposed formations. As an example, one or more of such factors may be taken into consideration by a framework such as the framework 600 (e.g., as inputs, models, etc.).

[0111] As an example, a mobile formation (e.g., salt or shale) can behave in a plastic manner. When compressed by overburden, material may flow and squeeze into a bore, thereby constricting or deforming the hole and trapping the tubulars. As an example, one or more of such factors may be taken into consideration by a framework such as the framework 600 (e.g., as inputs, models, etc.).

[0112] As an example, overpressured shales can be characterized by formation pore pressures that exceed normal hydrostatic pressure. Insufficient mud weight in these formations may permit a hole to become unstable and collapse around pipe. As an example, one or more of such factors may be taken into consideration by a framework such as the framework 600 (e.g., as inputs, models, etc.).

[0113] Reactive shales and clays tend to absorb water from drilling fluid. Over time — ranging from hours to days — they can swell into a bore. As an example, one or more of such factors may be taken into consideration by a framework.

[0114] As an example, drillstring vibration may cause caving of a bore. Such cavings can pack around a pipe, causing it to stick. Downhole vibration may be controlled by monitoring parameters such as weight on bit, rate of penetration and rotary speed, which can be adjusted from a driller’s console or, for example, via issuance of one or more signals from a framework. As an example, one or more of such factors may be taken into consideration by a framework.

[0115] As an example, differential sticking may happen when a drillstring is held against a bore by hydrostatic overbalance between bore pressure and pore pressure of a permeable formation. Such an issue may occur when a stationary or slow-moving drillstring contacts a permeable formation, and where a thick filtercake is present. As an example, a depleted reservoir may be a cause of differential sticking. As an example, one or more of such factors may be taken into consideration by a framework.

[0116] As an example, keyseating can take place when rotation of a drillpipe wears a groove into the borehole wall. When the drillstring is tripped, the bottomholeassembly (BHA) or larger-diameter tool joints can be pulled into the keyseat and become jammed. A keyseat may also form at the casing shoe if a groove is worn in the casing or the casing shoe splits. Such an issue can occur at abrupt changes in inclination or azimuth, for example, while pulling out of the hole and after sustained periods of drilling between wiper trips. Wireline logging tools and cables may be susceptible to keyseating. As an example, one or more of such factors may be taken into consideration by a framework.

[0117] As an example, an undergauge hole may develop while drilling hard, abrasive rock. As the rock wears away the bit and stabilizer, the bit drills an undergauge, or smaller than specified, hole. When a subsequent in-gauge bit is run, it can encounter resistance in the undergauge section of hole. If a string is run into a hole too quickly or without reaming, a bit can jam in the undergauge section. Such an issue may occur when running a new bit, after coring, while drilling abrasive formations, when a PDC bit is run after a roller cone bit, etc. As an example, one or more of such factors may be taken into consideration by a framework.

[0118] As an example, cement blocks may pack-off a drillstring, for example, when hard cement around a casing shoe breaks off and falls into a new openhole interval drilled out from under casing. Uncured, or green, cement may trap a drillstring after a casing job. For example, when the top of cement is encountered while tripping in the hole, a higher than expected pressure surge may be generated by a BHA, causing the cement to set instantaneously around the BHA. As an example, one or more of such factors may be taken into consideration by a framework.

[0119] As an example, a collapsed casing can occur when pressures exceed a casing collapse pressure rating or when casing wear or corrosion weakens the casing. The casing may also buckle as a result of aggressive running practices. Such conditions may be discovered when a BHA is run in the hole and hangs up inside the casing. As an example, one or more of such factors may be taken into consideration by a framework.

[0120] As an example, one or more hole cleaning problems may prevent solids from being transported out of a bore. When cuttings settle at the low side of deviated wellbores, they may form layered beds that may pack around a BHA. Cuttings and cavings may also slide down an annulus when pumps are turned off, thus packing around a drillstring. Such issues may occur due to one or more of low annular flowrates, inadequate mud properties, insufficient mechanical agitation and short circulation time. As an example, one or more of such factors may be taken into consideration by a framework.

[0121] As an example, a framework may include one or more features of a framework described in US 2018 / 0171774 A1 , published 21 June 2018, which is incorporated by reference herein in its entirety.

[0122] FIG. 9 shows an example of a framework 900 (e.g., a computational framework) that can be adapted to generate one or more workflows, for example, as indicated by various arrows. As shown in the example of FIG. 9, the framework 900 can receive information as input such as, for example, one or more rigstates (e.g., operational states, non-operational states, etc.), surface torque, mud cake thickness and / or quality, BHA stab type and / or slickness, wellbore inclination, information from one or more offset bores (e.g., wells), mud logs, real-time caliper measurements, rate of penetration, flow rate(s), etc.

[0123] As mentioned, a framework can be at least in part model-based. In FIG. 9, models can include a torque and drag (T&D) model comparison component 952, a hydraulics model 954, a geomechanics model 956, and optionally one or more other types of models, which may be selectable and utilized in a dynamic manner during execution of a workflow (e.g., adaptive execution of a model-based workflow).

[0124] In the example of FIG. 9, the framework 900 is shown as including various input components such as a pickup / slack-off (PU / SO) determination component 912, a stationary time determination component 914, and a sign(s) of borehole washout component 916. The framework 900 can include, for example, a torque spikes at rotation start component 922, a cavings detection / observation component 942, an amount of overbalance component 944, a stand pipe pressure (SPP) condition decision component 962, a wellbore stability decision component 964, a cuttings detection / observation component 966 (e.g., cutting carried to surface, etc.), and a mud balance component 968 (e.g., a decision or decision-making component as to mud / drilling fluid, etc.). In such a framework, the various components can be associated with one or more sources of information, which can include, for example, one or more sensors that are at one or more field sites where field operations may be performed.

[0125] As an example, as to mud balance, information pertaining to mud balance may be acquired via one or more sensors that can measure density (weight) of mud, cement or other liquid or slurry. As an example, one or more pressurized mud balance sensors may be utilized. As shown in FIG. 9, the mud balance component 968 can output information to the risk of solids-induced pack-off component 984.

[0126] As to output information, components of the framework 900 can include, for example, a potential for differential sticking component 972, a risk of getting differentially stuck component 982, and a risk of solids-induced pack-off component 984. The framework 900 may be extensible in that one or more components can be added, deleted, updated, etc. In such an example, the framework 900 may be adapted to particular types of equipment, fluid, energy sources, etc., at a site.

[0127] As mentioned, one component of the framework 900 may generate output that can be received by one or more other components of the framework 900. For example, the risk of solids-induced pack-off component 984 may receive output from the T&D model comparison component 952, from the hydraulic model component 954, from the wellbore stability decision component 964, the cutting detection / observation component 966, etc. In such an approach, the framework 900 provides for forward progression from input(s) to output(s) as well as, for example, backward progression where an output or outputs may be traced backwards to one or more reasons, causes, datum, data, source of data, sources of data, etc. In such an example, an identified cause (e.g., contributing cause, etc.) to an output (e.g., an indicator such as an alarm, etc.), may be utilized to generate one or more signals that can be issued from the framework 900 to one or more pieces of equipment (e.g., to effectuate control of such one or more pieces of equipment as to one or more field operations, etc.).

[0128] As an example, the framework 900 may be adaptable in real-time, for example, where caliper information becomes available, the signs of borehole washout component 916 may be utilized to provide output to the wellbore stability decision component 964. As another example, where mud log data indicates that cavings are observed per the component 942, information may be transmitted to and received by the wellbore stability component 964. In such an approach, the quality of information provided to the risk of solids-induced pack-off component 984 may be enhanced,which may increase accuracy of output of the risk of solids-induced pack-off component 984.

[0129] In the example of FIG. 9, the arrows in the framework 900 pertain to an example of a workflow, as may be enabled to assess conditions germane to stuck pipe. Inputs to the various components are illustrated along the left side in FIG. 9 while some outputs are illustrated along the right side in FIG. 9. As an example, inputs can include raw measurements, for example, acquired and / or made at a rigsite during real-time execution and / or, for example, acquired pre-job such as from one or more offset wells and / or, for example, from a well plan (e.g., and / or from one or more analyses made before drilling). In the framework 900, components toward the right edge can compute various types of risk such as, for example, a risk that a pipe and / or tool may be about to become stuck. In such an example, the framework 900 may issue one or more signals, which can include one or more of an alarm signal, a control signal or another type of signal. As an example, the framework 900 may progress backwards from an output such as an alarm to one or more likely contributing causes of the alarm (e.g., one or more components of the framework 900 and associated input, etc.). In such an example, the framework 900 may issue one or more recommendations (e.g., one or more recommended actions) and / or one or more control signals, which may, for example, be issued via one or more network interfaces addressed to one or more pieces of equipment at a rigsite for control of at least one of the one or more pieces of equipment at the rigsite.

[0130] In the example of FIG. 9, the framework 900 can provide for combinations of various indicators, which may include those that have been tried in the past (e.g., with or without success). Various indicators may be one or more of data-driven, physics-based model-driven, etc.

[0131] As an example, the framework 900 can include estimating uncertainty of one or more variables and optionally propagating such uncertainty from one component to another component (e.g., in a chain or chains). Such an approach can allow for different measurements of similar quantities to be merged. For example, the two components 982 and 984 that compute risk of stuck pipe can take multiple indicators of stuck pipe as inputs and generate a single probability of risk.

[0132] As an example, the components 982 and 984 can be part of a Bayesian network, which can optimally consider uncertainty of each input to compute risk. Insuch an example, the Bayesian network can allow the computation of risk in a manner that differs from a “black-box” approach. For example, when the risk of stuck pipe is high, a user can interrogate the framework 900 to estimate which of the inputs are generating the high risk and why. Such an interrogation can include performing a backwards progression through the various components to identify an input or inputs, which, as mentioned, may be a basis for issuing one or more signals as to control of one or more pieces of field equipment (e.g., rigsite equipment, whether surface, subsurface, etc.).

[0133] As an example, a Bayesian network can include weights where the weights are associated with data acquired from equipment such as equipment in a field that performs one or more field operations. For example, a Bayesian network can include weights that are applied to data-based numbers where the data are acquired from equipment at a rigsite, which can include surface and downhole equipment.

[0134] As an example, a Bayesian network may be a type of graphical model that uses probability related to one or more types of events. Such a network may be a belief network or a causal network. As an example, a probabilistic network may include one or more directed cyclic graphs (DCGs) and a table of conditional probabilities to determine the probability of an event occurring.

[0135] In terms of an arc of a graph of a network (e.g., directed acyclic graph (DAG), etc.), an individual arc may have a weight or value associated with it, indicating a strength of interaction between nodes that the arc connects. The nature of such a weight can be application dependent. For example, it may represent a cost associated with a particular action, the strength of a connection between two nodes or, in the case of probabilistic models, the probability that a particular event will occur. As an example, a Bayesian belief network (e.g., a Bayesian network) can be conducive to understanding a scenario or scenarios as they can be constructed such that a parent(s) of a variable can be a direct cause. Such an approach can help to facilitate a process of determining weights for arcs that connect nodes of a network (e.g., assessment of conditional probabilities, etc.).

[0136] As an example, a Bayesian network can be implemented as part of a computational framework that includes one or more interfaces (e.g., one or more network interfaces, etc.) that can receive data acquired at one or more sites such asone or more rigsites. As an example, a computational framework can include one or more processors, memory, interfaces, etc. As mentioned, a computational framework can include receiving data, which may include sensor data from one or more sensors. As an example, a computational framework can provide for sensor fusion utilizing at least in part a Bayesian network (e.g., or Bayesian networks).

[0137] Sensor fusion refers to the class of problems where data from various sources can be integrated to arrive at an interpretation of a situation (e.g., a scenario). For example, data from various rigsite sensors, which may be for different sampling rates, different data formats, different units, etc., can be integrated to determine a status of one or more rigsite operations, which may include one or more operations that are associated with sticking (e.g., actual sticking, breaking free or becoming unstuck, risk of sticking, likelihood of success of breaking free, etc.). As an example, a sensor fusion approach may include receiving data from a plurality of sensors where a state can be discerned for a system by integrating at least a portion of the received data.

[0138] As an example, a computational framework that implements one or more Bayesian networks may provide for handling of data that may differ as to one or more of temporal resolution and spatial resolutions. In such an example, the framework may solve a so-called correspondence problem that can include deciding which event(s) from one sensor correspond to the same event(s) as reported in one or more other sensors. As an example, a Bayesian network approach can be implemented in a manner that tends to be robust to missing data via combining data from a plurality of sources (e.g., a plurality of sensors, etc.). As an example, a Bayesian network approach may address scenarios where each sensor individually may have a limited chance of providing an acceptable interpretation by combining information from a plurality of sensors to increase the likelihood of providing an acceptable interpretation (e.g., valid in terms of physics, matching real-world conditions, etc.).

[0139] As an example, a computational framework that implements one or more Bayesian networks can provide for diagnosing one or more issues that a system may be experiencing, may have experienced, may likely experience, may experience given one or more sets of operational conditions, etc. As an example, a computational framework may provide for deciding when to issue a signal, which may be a control signal and alert signal, etc.

[0140] As an example, parameters of a Bayesian network may be tuned as conditional probability tables, which may be relative weights of the Bayesian network. In such an example, data can be used to tune parameters where the parameters have physical meaning as they refer to input indicators. In such an example, where output of a computational framework that implements at least one Bayesian network indicates that a risk for a sticking or other issue at a rigsite is high (e.g., compared to a threshold), as various parameters are associated with real-world data acquired at the rigsite, a method can include working backwards through a Bayesian network to help identify one or more causes of the indicated risk where such a method can include issuing one or more signals that aim to address the one or more causes. For example, where an indicated risk is for differential sticking of a drillstring in a borehole that is being drilled, working backwards may identify that a hole cleaning index (HOI) may be a cause where the hole cleaning index is based at least in part on a flow rate for drilling fluid (e.g., mud) such that a signal may be issued to control a pump or pumps that pump drilling fluid to, for example, increase the flow rate for the drilling fluid. In such an example, where the increase in the flow rate for the drilling fluid does not adequately address the indicated risk (e.g., sticking occurs), one or more weights may be adjusted as to flow rate for the drilling fluid on the hole cleaning index or, for example, a weight for the hole cleaning index may be adjusted where another cause may be contributing, at least in part, to the indicated risk.

[0141] As an example, a computational framework that includes a Bayesian network or networks can provide for backward progressions from an outcome (e.g., an indicator) to data and / or one or more sources of data. Such an approach can help to associate the outcome with particular, specific data and / or one or more specific sources of data (e.g., equipment, acquisition equipment, etc.). As to specific data, as mentioned, data may be temporal and / or spatial with respect to resolution such that a backward progression may provide for identifying where temporal and / or spatial aspects of data can be improved. In such an example, a signal may be issued by a computational framework that controls one or more temporal and / or spatial aspects of a sensor or sensors that acquire particular data. Such an approach can help the computational framework to operate with an enhanced ability to provide more beneficial outputs (e.g, operate with greater certainty, both in a forward mode and in a backward mode, etc.).

[0142] As an example, a computational framework can be extensible such that one or more components can be added, removed, updated, etc. For example, where the computational framework includes a Bayesian network as to risk of differential sticking and where a new piece of equipment is implemented at a rigsite that outputs data, the Bayesian network may be augmented via instantiation of a component that adds the data as an input, which may be appropriately weighted, etc.

[0143] As an example, a Bayesian network can provide for computation of “potential risk” (e.g., for the case of differential sticking, see the component 972). Such an approach can consider the environment or set-up to determine prior risk. For example, a workflow may consider that certain BHAs are more susceptible to differential sticking, or that the chance of becoming differentially stuck increases with overbalance pressure (see, e.g., the component 944). Potential risk can be computed for a given environment and well plan, for example, before drilling commences and may be used to help design the well plan. As an example, the framework may provide output that can assist in choosing a BHA and / or mud weight that minimizes potential for differential sticking. Such an approach may utilize a component or components that may be part of a Bayesian network and / or a physics-based model (e.g., of a process of differential sticking).

[0144] As an example, one or more outputs from the framework 900 may be received as one or more inputs to a system that can generate information suitable for rendering to a display (e.g., instructions to control a graphics processing unit, etc.). For example, where a BHA is being designed, information may be output to a well plan GUI. In such an example, a drillstring GUI may highlight a portion of the drillstring graphic that can be selected based at least in part on output from the framework 900. For example, where the framework 900 indicates a risk of getting stuck at a certain depth during drilling operation, a tool may be highlighted and information rendered to the display as to the type of sticking event, the risk of the sticking event occurring, the conditions and / or causes of the sticking event (e.g., and associated risk and / or uncertainty and / or contribution to the cause or causes) and, for example, a recommended change or adaption to the drillstring, whether physically or operationally that may be implemented in an effort to reduce the risk of getting stuck.

[0145] As an example, an output from a framework such as the framework 900 may include a suggested alteration to a well trajectory and / or to a completion of a well.As an example, an output from the framework 900 may provide for highlighting one or more points in a point spreadsheet as to one or more parameters of a well or wells. As an example, a user may accept a change and / or otherwise adapt a well plan based at least in part on output from the framework 900. For example, a user may utilize a GUI to edit one or more portions of a well plan, optionally based at least in part on information communicated with one or more members of a team.

[0146] As illustrated in the example framework 900 of FIG. 9, which includes arrows as to an example of a workflow, risk of getting stuck can be based on two mechanisms: risk of differential sticking per the component 982 and risk of solids- induced pack-off per the component 984. As mentioned, these components can be part of a Bayesian network. Such a network may be extended to compute the risk of getting stuck based on one or more other mechanisms (e.g., key seating, shale swelling, etc.), alternatively or additionally.

[0147] As to various operations, conditions, etc., associated with sticking, consider hole cleaning, hydraulics, and torque and drag as factors that can be considered. As shown in FIG. 9, the framework 900 can include components associated with such factors (see, e.g., the component 954 as to hydraulics and hole cleaning index (HCI), the components 962 and 952 as to torque and / or drag, etc.).

[0148] As an example, a method can include one or more of assessing stand pipe pressure (SPP) actual versus model (e.g., monitor separation of curves and determine whether a result is to be part of a workflow, etc.; see also the component 962 of the framework 900), assessing equivalent circulating density (ECD) actual versus model (e.g., monitor separation of values and determine whether a result is to be part of a workflow, etc.), assessing torque and drag (T&D) actual versus model (e.g., monitor for deviation of trend and model and determine whether a result is to be part of a workflow, etc.; see also the component 952 of the framework 900).

[0149] As an example, a service may provide for analysis of real-time data streams from a number of sensors to provide real-time status of a well, for example, by combining depth- and time-based data related to wellbore pressure balance, drilling mechanics, and / or hole condition.

[0150] As an example, a service may provide wellbore pressure balance for detection of fluid influx or loss in the wellbore, even at very low volumes. Early kickdetection may be supported by pump-off gas analysis to identify potential underbalance situations. An automatic flowback fingerprint may be captured.

[0151] As an example, a service may provide for drilling mechanics analysis such as wear and behavior of a drill bit, which may be assessed by monitoring one or more drilling parameters (e.g., axial (bit bouncing) and torsional (stick / slip) vibration frequencies and energy) through measurements made by one or more sensors. Potential issues may include bit balling, drillstring vibration, and bit wear, which may be predicted as to risk and risk potential where one or more drilling parameters may be optimized to improve ROP and increase equipment life.

[0152] As an example, a service may provide for assessing hole condition such as wellbore stability and hole-cleaning efficiency, which may be analyzed in real-time by comparing measurements of relevant parameters (e.g., pickup, slack-off, and free- rotating weights; torque; and equivalent circulating density (ECD)), with theoretical values calculated using one or more models. As an example, hole-condition monitoring can be linked with data from one or more cuttings flowmeters.

[0153] As an example, one or more of a hydraulics physics-based model, a geomechanics physics-based model, or other type of physics-based model may be implemented as part of a framework, for example, to perform one or more workflows. As an example, the framework 900 of FIG. 9 may be operatively coupled to a framework such as the PETREL framework and / or the DELFI framework environment (SLB, Houston, Texas).

[0154] As an example, a computational framework may output information with respect to one or more operational parameters as to one or more operators, one or more service providers, one or more suppliers, etc. As an example, a computational framework may output information that associates decision making and one or more operators (e.g., individual, team, etc.). Such an approach may help to identify how decisions are made during operations in the field. Such an approach may help to assess decision making by an individual, a team, etc., which may provide for tuning of one or more parameters of a computational framework (e.g., one or more Bayesian network parameters, etc.).

[0155] As an example, a computational framework can respond to real-time decision making in the field during operations. As a Bayesian network can include inputs as to acquired, real-time data, outputs of the Bayesian network can depend onthe acquired, real-time data. As an example, a computational framework may associate changes in real-time data with real-time decision making by one or more operators, controllers, etc. As an example, an output of the computational framework can include issuing a query to one or more onsite devices such as, for example, an operations computer, etc. As an example, consider issuing a query to a device that asks “was the mud flow rate increased?” In such an example, a response can be received by the computational framework, which may be utilized to adjust one or more parameters (e.g., one or more weights, etc.).

[0156] As an example, a framework may include one or more features of a system as described in US Patent No. 7,861 ,800, issued 4 January 2011 , which is incorporated by reference herein.

[0157] FIG. 10 shows an example of a flow diagram depicting the flow of information through various components is shown according to an illustrative embodiment. In FIG. 10, a data processing system 1010 is shown that may provide for handling of decision factors 1012 that may be entered into a belief network generator 1014 (e.g., a Bayesian belief network generator, etc.). Such decision factors 1012 may be a set of causal variables that are considered when arriving at a conclusion. A causal variable may be a factor that might be considered when arriving at a conclusion. A set as used herein may refer to one or more items. For example, a set of causal variables are one or more causal variables

[0158] The decision factors 1012 can also be those conclusions that can be ascertained from the set of causal variables. The decision factors 1012 generally relate to a condition encountered in a drilling operation, and a remedial action that can be performed in response to that condition. The decision factors 1012 may be obtained from surveys, questionnaires, data logs, or other sources of information.

[0159] The decision factors 1012 may be entered into the belief network generator 1014. As an example, the belief network generator 1014 may be implemented using a process executing on the data processing system 1010. The belief network generator 1014 may assign each of the decision factors 1012 that are entered into a node. As an example, the belief network generator 1012 may then causally associate the generated nodes to form a belief network 1016.

[0160] As an example, the belief network 1016 may be used to compensate for inherent uncertainty in knowledge-based applications such as well sitetroubleshooting. As an example, the belief network 1016 may be a problem, or set of problems, that may be modeled as a set of nodes interconnected with pathways to form a directed acyclic graph (DAG). In such an example, each node within the belief network 1016 represents a random variable, or uncertain quantity, which can take two or more possible values. Pathways may signify the existence of direct influences between the linked variables.

[0161] The various nodes of a belief network may be associated in a cause / effect arrangement. For example, each node containing a causal variable input, a causal node, from the decision factors 1012 may be located upstream from a conclusion. By weighing the various upstream nodes, the belief network 1016 may be able to generate a conclusion from those nodes. A conclusion node can be a node of the belief network 1016 that includes a conclusion generated from the weighting of the associated causal nodes.

[0162] Conclusion nodes themselves may be causal nodes for a subsequent downstream conclusion. For example, a first node is a conclusion node for a set of causal nodes. That first node may itself be one of a second set of causal nodes that feed into a second node, the second node being a conclusion node for the second set of causal nodes.

[0163] Various nodes of a belief network may be connected using an interactive template having a graphical user interface (GUI). The interactive template can present the user with the set of nodes and allow the user to connect the nodes in the desired fashion. A user can then associate the nodes in a desired fashion to create the desired cause / effect relationship between the various nodes of the belief network 1016. As an example, a GUI may be one illustrated method of associating the various nodes to create the belief network 1016. For example, consider one or more other methods, such as, for example, a parse of the decision factors 1012, language recognition of the decision factors 1012, or other methods of classifying and connecting various ones of the decision factors 1012 input into the belief network generator 1014.

[0164] Once the belief network generator 1014 has generated the belief network 1016 from the decision factors 1012, the belief network generator 1014 may forward the belief network 1016 to a multinet builder 1018. The multinet builder 1018 may be implemented using instructions executed by the data processing system 1010,which may connect common nodes of separate belief networks, such as the belief network 1016, to form the multinet belief network 1020.

[0165] When the multinet builder 1018 receives the belief network 1016 from the belief network generator 1014, the multinet builder 1018 parses the belief network 1016 to determine contents of each node contained therein. The contents of the nodes may be the decision factors 1012 that were entered into the belief network generator 1014.

[0166] The multinet builder 1018 may identify a current version of the multinet 1020 from an associated data storage 1022. The multinet 1020 may be a combination of previous separate belief networks, such as, for example, the belief network 1016. Common nodes among the separate belief networks may be associated. As an example, common nodes can be nodes of different belief networks, or nodes contained in a current multinet iteration, that contain identical or substantially similar decision factors, such as one or more of the decision factors 1012.

[0167] The multinet builder 1018 may parse the multinet 620 to determine the contents of each node contained therein. The contents of the nodes are those decision factors, such as the decision factors 1012 that are previously entered into and incorporated into a belief network 1016 by belief network generator 1014.

[0168] The multinet builder 1018 may then compare the parsed nodes from the belief network 1016 and the parsed nodes from the multinet 1020 to identify common nodes. Common nodes among the separate belief networks may be associated. Common nodes may be nodes of different belief networks, or nodes contained in a current multinet iteration, that contain identical or substantially similar decision factors, such as one or more of the decision factors 1012. Common nodes may also be identified and connected manually by a user utilizing a graphical user interface.

[0169] Once common nodes between the belief network 1016 and the multinet 1020 are identified, the multinet builder 1018 may then associate the generated nodes to create an updated version of the multinet 1020. Common nodes in the belief network 1016 and the multinet 1020 may be overlapped, so that conclusion nodes of the belief network 1016 and the multinet 1020 can be affected by each other’s causal nodes. In this manner, separate belief networks having separate or different causal nodes can be effectively combined into a multinet. Further, causal nodes that mayaffect more than one decision node can be combined into a unified decision model contained in the multinet 1020.

[0170] When common nodes have been associated, the multinet builder 1018 may save the updated version of the multinet 1020 to the associated data storage 1022. The multinet 1020 may then be available as a diagnostic or predictive analysis tool for generating parameters in response to a query by an operator or engineer.

[0171] As an example, the multinet answer product 1024 may receive iteration parameters 1026 from an operator or engineer. The multinet answer product 1024 may be implemented via instructions executed by the data processing system 1010. The iteration parameters 1026 may be observed conditions that relate to one or more conclusions. The iteration parameters 1026 may correspond to at least one decision factor, such as one or more of the decision factors 1012, as may be contained within a node of the multinet 1020.

[0172] Responsive to receiving the iteration parameters 1026, the multinet answer product 1024 may identify the multinet 1020 from the storage 1022. The multinet answer product 1024 may then input the iteration parameters 1026 into the corresponding nodes of the multinet 620 to generate oilfield parameters 1028. The oilfield parameters 1028 may be or include, typically, those conclusions from conclusion nodes in the multinet 1020.

[0173] The multinet answer product 1024 may then forward the oilfield parameters 1028 to an operator and / or a machine, whereby one or more of the oilfield parameters 1028 may be utilized in performing one or more drilling operations at a wellsite (e.g., a rigsite).

[0174] FIG. 11 shows an example of a framework 1100 that may be implemented for reducing risks of pipe sticking. As shown, the framework 1100 may be referred to as a stuck pipe averter. The framework 1100 may include one or more types of networks, which may be or include one or more belief networks (e.g., one or more Bayesian belief networks).

[0175] In the example of FIG. 11 , the framework 1100 may operate responsive to answers to questions whereby a belief network may generate conditional probabilities. For example, consider outputs such as one or more risk category scores, answers to one or more “what if” scenarios, identification of one or more root causes (e.g., from one or more root cause analyses, etc.).

[0176] As shown in FIG. 11 , a framework 1160 may be utilized where, for example, expert judgment may be relied upon to generate answers to a questionnaire for use in an explainable artificial intelligence (XAI) semantic causal network to generate conditional probabilities for an XAI report.

[0177] As an example, the framework 1100 and / or the framework 1160 may include one or more features of the framework 900 of FIG. 9 and / or one or more features of the framework 1000 of FIG. 10.

[0178] As an example, a system may provide for generation of enhanced responses for stuck pipe assessments. For example, a system may include one or more generative artificial intelligence (Al) components for generation of enhanced questionnaire responses for assisting stuck pipe risk analysis. For example, a system may include utilization of one or more large language models (LLMs).

[0179] A large language model (LLM) may be a type of language model notable for its ability to achieve general-purpose language understanding and generation. An LLM may acquire abilities by using relatively massive amounts of data to learn parameters (e.g., determine parameter values, etc.) during training. An LLM may be an artificial neural network or networks (e.g., consider a transformer, etc.) and may be trained and / or pre-trained using one or more types of learning (e.g., self-supervised learning, semi-supervised learning, unsupervised learning, etc.).

[0180] As an example, an autoregressive language model (e.g., AR LLM) may operate by taking input text and repeatedly predicting a next token or word. As an example, an LLM may be tuned, for example, for a particular domain. As an example, an LLM such as the Generative Pretrained Transformer (GPT) 3 (GPT-3) may be prompt-engineered. As an example, an LLM may acquire embodied knowledge about syntax, semantics and ontology inherent in human language corpora; noting that an LLM may also acquire inaccuracies and biases present in one or more corpora.

[0181] FIG. 12 shows an example architecture 1200 of a GPT; noting that one or more features of the architecture 1200 may be utilized in an LLM. As shown, the architecture 1200 may include various components that may be represented as block, etc. For example, consider matmul (e.g., matrix multiplication), mask, softmax, dropout, linear, Gaussian error linear unit (GELU), etc. As an example, a foundational GPT model may be further adapted to produce more targeted systems directed to specific tasks and / or subject-matter domains. Techniques for such adaptation mayinclude additional fine-tuning (e.g., beyond tuning of a foundation model, etc.), certain forms of prompt engineering, etc. As an example, an LLM may be a chatbot type of LLM. For example, consider the OpenAI ChatGPT LLM, which is an online chat interface powered by an instruction-tuned language model trained in a similar fashion to InstructGPT. Other chatbots may include features of GPT-4 (OpenAI), Bard (e.g., LaMDA family of conversation-trained language models, PaLM, etc.) (Google, Mountain View, California), etc.

[0182] As an example, a LLM Meta Al (LLaMA) LLM may be utilized, which includes a transformer architecture; noting some architectural differences compared to GPT-3. For example, LLaMA utilizes the SwiGLU activation function rather than ReLU, uses rotary positional embeddings rather than absolute positional embedding, and uses root-mean-squared layer-normalization rather than standard layernormalization. Further, there may be an increase in context length from 2K (Llama 1 ) tokens to 4K (Llama 2) tokens between.

[0183] As an example, a system may implement a Retrieval-Augmented Generation (RAG) approach (e.g., a Retrieval Augmented Generator) that may provide an LLM with additional information from an external knowledge source. In such an example, the LLM may be able to generate more accurate and contextual answers while reducing hallucinations.

[0184] FIG. 13 shows an example of a RAG architecture 1300 that includes a vector database that can generate context for a query to formulate a prompt for an LLM that can generate a response to the prompt. As shown, actions may include retrieve 1310, augment 1320, and generate 1330. As shown, the vector database may be utilized to augment a query such that the query may include context that may be beneficial in generation of output from the LLM. As an example, one or more features described in an article by Lewis et al. may be utilized (see “Retrieval- augmented generation for knowledge-intensive NLP tasks”, Advances in Neural Information Processing Systems, 33, 9459-9474 (2020), which is incorporated by reference herein in its entirety).

[0185] LLMs are trained on large amounts of data to achieve a broad spectrum of knowledge, which may be stored in neural network weights (e.g., parametric memory). However, prompting an LLM to generate a completion that demands knowledge that was not included in its training data, such as newer, proprietary, ordomain-specific information, can lead to factual inaccuracies (e.g., hallucinations). For example, consider the following scenario concerning ChatGPT:ChatGPT answering the question “What did the president say about Justice Breyer” with “I don’t know because I don’t have access to real-time information” ChatGPT’s answer to the question, “What did the president say about Justice Breyer?”

[0186] As seen in this scenario, there is a lack of a bridge across a knowledge gap between the LLM knowledge and additional context to help the LLM generate more accurate and contextual completions while reducing risks of hallucinations.

[0187] Neural networks tend to be adapted to domain-specific or proprietary information by model fine-tuning. While fine-tuning may be helpful (e.g., effective), fine-tuning introduces computational demands, which may be intensive, expensive, and requiring technical expertise, which may diminish agility to adapt to evolving information.

[0188] A RAG architecture provides for more flexibility through combination of a generative model with a retriever component to provide additional information, for example, from one or more external knowledge sources, which may be, in various instances, updated more readily.

[0189] A RAG architecture has been likened to an open-book exam, as additional information can be provided to an LLM such that the LLM generates a “better” response to a query. As an example, factual knowledge may be effectively separated from an LLM’s reasoning capability and stored in an external knowledge source, which may be readily accessed and, as appropriate, updated. As an example, a RAG architecture may provide for parameter knowledge (e.g., learned during training that is implicitly stored in neural network weights) and non-parametric knowledge (e.g., stored in an external knowledge source, such as a vector database).

[0190] As shown in the example of FIG. 13, the system 1300 may implement a RAG workflow from query through retrieval with a vector database to prompt stuffing and finally response generation. As shown, as to the retrieve component (e.g., retrieval component), a query is used to retrieve relevant context from an external knowledge source. For example, a query may be embedded with an embeddingmodel into the same vector space as the additional context in the vector database. In such an example, a similarity search may be performed where a top number of closest data objects from the vector database may be returned. As shown, output from the vector database similarity search may be utilized to augment a prompt. For example, the query and the retrieved additional context may be stuffed into a prompt template. Next, the prompt, as a retrieval-augmented prompt, may be fed to an LLM where the LLM generates a response.

[0191] FIG. 14 shows an example of a system 1400 that may provide for generation of output utilizing a generative process, which may include implementing one or more large language models (LLMs) that can, responsive to input, generate output. As shown, the system 1400 can include a vector database (Vector DB) component, a retriever component and an LLM component.

[0192] As an example, a vector database may be a type of specialized database designed to store and retrieve vector embeddings, which may be present in the form of numerical arrays representing various characteristics of an object. Such embeddings may be distilled representations of training data and / or other data, for example, serving as a filter through which new data may be run during an inference part of a machine learning process. As explained, in an RAG architecture, vector embeddings may provide for query augmentation, for example, to augment a query with contextual information.

[0193] As an example, a vector database may be operatively coupled to a LLM via a retriever component. In the context of LLMs, vector databases can store the vector embeddings, which may include those resulting from model training and / or one or more other sources. In such an example, performance of database-based similarity searches may be improved, where an aim may be to find a best match between a prompt and a particular vector embedding.

[0194] Vector embeddings may be considered to be numerical representations of data objects. Vector embeddings may be generated by one or more processes and serve as a distilled, structured representation of data. As an example, each point in a high-dimensional space may provide for a correspondence to a unique data object where distance between points represents similarity between the corresponding data objects. In the context of vector databases, vector embeddings may be used to transform and store data in a way that allows for efficient similarity search, which maybe utilized in applications such as semantic search and natural language processing (NLP), where a goal may be to find one or more data objects that are semantically similar to a given query and / or otherwise relevant to a given query.

[0195] As an example, the system 1400 may employ vector indexing. For example, once data has been transformed into vector embeddings, these embeddings may be stored in a manner that allows for efficient search and retrieval. Vector indexing involves organizing and storing vector embeddings in a database in a way that allows for efficient similarity search. The high dimensionality of a vector space and demands to perform complex distance computations can present challenges where such challenges may be suitably handled using advanced indexing algorithms and data structures. Vector databases tend to handle such tasks efficiently.

[0196] As to similarity searches in vector databases, the concept of similarity search in vector databases can involve finding the most similar vectors to a given vector within the database. Similarity may be determined, for example, using one or more distance metrics (e.g., Euclidean distance, cosine similarity, etc.). In the context of LLMs, a similarity search may be used to find a best match between a prompt and a stored vector embedding, which may allow an LLM to generate appropriate responses to questions based on its training (e.g., training data) where, as explained, questions may be augmented (e.g., through a vector database-based search result). As explained, a vector database may be provided as a front end to an LLM whereby one or more searches may generate information that may augment a query as in an RAG architecture.

[0197] As shown in the example of FIG. 14, the system 1400 may provide for transforming data from one or more sources (e.g., source document library, etc.) to vectors for storage in the vector database component. As shown, data may include image data, tabular data, text data, etc. As an example, such data may include training data that were utilized to train the LLM. As an example, such data may include data from a rigsite, which may be streamed data such as, for example, real-time streamed data. As an example, a vector database may be dynamically maintained to provide for up-to-date, timely information that may be utilized to augment or otherwise enhance a query.

[0198] As shown in the example of FIG. 14, the system 1400 may utilize the retriever component as an interactive component that interacts with the vectordatabase and that provides output to the LLM. For example, the retriever component may receive one or more questions (e.g., manual, automatic, batch, etc.) and provide for generation of a prompt to the LLM whereby the LLM may generate one or more responses (e.g., answers) responsive to receipt of the prompt. As shown, the retriever component may generate a prompt that is dependent upon a question (e.g., a query) and a context instruction (e.g., contextual information extracted from the vector database using a similarity and / or other type of search). In such an example, a question may be presented with context to the LLM whereby the LLM may generate one or more responses thereto.

[0199] In the example of FIG. 14, the output of the system 1400 may be provided to one or more frameworks, which may include, for example, one or more stuck pipe assessment frameworks. As shown, the output may be considered “expert judgment”, which may be based on machine learning techniques applied to the source data, which may be utilized for training one or more machine learning models.

[0200] The system 1400 of FIG. 14 may provide for addressing difficulties of discovering relevant observations related to borehole condition from a body of source documents such as, for example, daily drilling reports, geological operations reports, well planning documents, end of well summaries, etc. As explained, extracted observations may be effectively passed to a questionnaire that generates a Bayesian conditional probability assessment and root cause analysis.

[0201] The system 1400 of FIG. 14 may provide for connecting a knowledge system (e.g., source documents, document parsing methods, vector database, retrieval augmented generation methods, and specialized large language model prompting) with one or more stuck pipe risk assessment frameworks.

[0202] As an example, a system may operate in an automated and / or semiautomated manner. Such an approach may conserve time and provide for more relevant output, which may also be more informed, more consistent, etc. Such an approach may be in contrast to manual reviews of tens to hundreds of source documents when looking for relevant observations related to stuck pipe analysis. Such manual searches tend to be prone to representativeness and interpretation biases, which may impair completeness and quality of subsequent stuck pipe analyses.

[0203] As an example, a system may implement Gen-AI technology to return context-relevant observations from a body of diverse source documents. As an example, a drilling domain chatbot operatively coupled to a stuck pipe analysis framework may provide for improved and / or more timely generation of conditional probabilities and root cause analysis.

[0204] As an example, an RAG-based system may enhance an existing analysis tool to understand root cause and actionable mitigation plans of stuck pipe, for example, by extracting relevant observations from one or more offset wells and / or one or more analogous hole sections. As an example, an RAG-based system may help to mitigate bias issues associated with manual information search and retrieval, for example, by using curated drilling domain expert prompts to retrieve information from source documents in near real-time. As an example, an RAG-based system may be used for pre-drill well planning and / or to modify operational plans prior to running casing or tripping, for example.

[0205] In evaluating or predicting risk of stuck pipe in drilling activity, a workflow may include acquiring data that may include, for example, data at a variety of decisionmaking points. The data to be acquired may include measurable parameters and controllable drilling practices (e.g., torque spikes while drilling, loss circulation while drilling, POOH, BHA change, etc.). As an example, one or more drilling engineers may operate within a workflow to compare a current well with one or more historical wells (e.g., in a common hole section, etc.) to better understand a situation and to make a decision whether one or more mitigation operations exist that may be implemented. As an example, in various instances, a system may operate in an automated manner whereby output of a system may be utilized to control one or more field operations.

[0206] For each well, drilling engineers normally need to review a variety of documents (e.g., daily drilling reports, end of well reports, and drilling data, etc.) that may be available and fill a list of questionnaires with answers that may then be sent to one or more stuck pipe prediction tools. In such a manual approach, acquiring and reviewing information from a large number of reports with different types and layouts demands a substantial amount of human effort, which is time-consuming and may be prone to errors, as may be due at least in part to complexity and / or volume of data involved.

[0207] As explained, to improve the efficiency of extracting relevant information from reports, a system may include a question-answering chatbot with integration of an LLM (e.g., consider an OpenAI GPT, etc.) and RAG components for query enhancement. As an example, an LLM may be trained on a vast amount of text data (e.g., billions of parameters to learn patterns existing in text data). As an example, an LLM may utilize a transformer architecture, whose self-attention mechanisms may help to better understand context, semantics, and structure of text. In such an example, the LLM may more accurately extract and summarize information from text documents.

[0208] As an example, a system may include one or more vector databases. For example, consider creation of a vector database by using an OpenAI embedding model that may be implemented to convert documents into vector representatives. In such an approach, the documents may be or include various types of data (e.g., image, tabular, text, sensor, etc.) for a well or wells. For example, consider an approach that generates one or more vector databases for a well where one or more field operations are being performed that may have associated risks such as, for example, stuck pipe risks. As an example, a dynamic approach to vector database management may be implemented, which may provide for contextual enhancement of queries submitted to one or more LLMs, which, in turn, may generate output relevant to stuck pipe risks.

[0209] As explained, an RAG architecture may be implemented where a retriever component may retrieve a portion of documents / contexts that relevant to a query or queries (e.g., a questionnaire or questionnaires), where the retriever component may interact with one or more vector databases. Such an approach may provide for query enhancement for an LLM, which may generate answers to enhanced queries. As an example, an LLM may be fed in a manner such that the LLM may generate one or more answers to a questionnaire based on retrieved context, which may help to ensure that such one or more answers are grounded to provided well documents, etc. As an example, a system may allow for users to conduct prompt engineering using a chatbot interface, which may provide for more relevant explanations to help an LLM to understand better a task and to locate appropriate information. For example, consider a query or questionnaire that may ask: “any note of POOH?”. In such an example, a user may supplement the query or questionnaire by adding information such as, for example, a definition or explanation of “POOH”:“POOH stands for pull out of hole”. As an example, one or more RAG components may provide for generation of such a definition or explanation automatically upon receipt of a query or questionnaire.

[0210] As explained, the system 1400 of FIG. 14 may be operatively coupled to a framework or frameworks for analyzing one or more types of issues, conditions, etc., such as, for example, stuck pipe. In such an approach a stuck pipe averter framework may be imparted expert knowledge via at least an LLM, which may respond to queries and / or enhanced queries. As explained, one or more RAG components may provide for tailored enhancement of queries, which may be within a domain germane to a well, particular operations performed at the well, etc.

[0211] As an example, a system may be run programmatically, passing curated expert questions to a retriever, whether individually and / or in a batch mode, and / or manually through a chatbot interface. Such a system may retrieve context-relevant drilling observations supported with citation from an original source document library, which may enable an expert to validate responses that feed into a stuck pipe analysis framework. As explained, a system may operate with or without a human-in-the-loop (HITL). For example, a system may operate in an automated and / or a semi-automated manner to improve risks assessment and / or operational control of field operations at a rigsite (e.g., at a wellsite where a well borehole is being drilled).

[0212] FIG. 15 shows an example of a system 1500 that includes RAG components operatively coupled to an LLM operatively, which are operatively coupled to a stuck pipe averter framework. Additionally, output of the stuck pipe averter framework may be processed to generate semantic explainable Al (XAI) as to risk ontology, which, in turn, may be fed to an LLM where conditional probabilities as output by the stuck pipe averter framework may be fed to an Al generated comparator component. As shown, output of the LLM may provide for comparing a scenario identified (e.g., via a multi-criteria search, etc.), for causal XAI validation. As shown, output of the Al generated comparator component may be fed to a data store that may receive queries (e.g., questionnaires) and risk ontology information, which may be available via interactions with a vector database. In such an example, the retriever component of the RAG architecture may receive information from the data store for purposes of query enhancement.

[0213] As an example, a system may provide for single well (e.g., single rigsite) and / or multiple wells (e.g., multiple rigsites). As an example, where multiple wellsites are involved, a batch of queries (e.g., questionnaires) may be generated where individual queries may be broken out and processed individually such that enhancement (e.g., augmentation) may be appropriately tailored for each of the individual queries.

[0214] As an example, a system may operate on the basis of processing drilling reports. For example, consider a system that may process a drilling report, which may be a background process, to generate an enhanced drilling report, which may include annotations and / or other information as to scenarios encountered during drilling.

[0215] As an example, a system may provide for generating input to a probabilistic framework. Such a system may provide for knowledge extraction in a query-response type of approach where a response may be fed to a probabilistic framework. In such an example, the probabilistic framework may utilize a probabilistic graph type structure (e.g., one or more decision trees, etc.) where output from such a structure may be improved by improvement of input (e.g., enhancement of prompts to an LLM such that the LLM generates output that may provide for improved operation of a probabilistic framework).

[0216] In the example of FIG. 15, the stuck pipe averter may be a computational framework (e.g., a pipe stuck averter framework) that is improved through use of a chat type of computational framework (e.g., a chat framework). For example, consider the questionnaire of the stuck pipe averter as being a series of questions that may be posed to a number of individuals that respond to the series of questions such that the stuck pipe averter framework may generate an answer or answers. In such an example, the number of individuals may be assisted by the chat framework. For example, the series of questions (e.g., consider a batch of questions) may be submitted to the chat framework as prompts that may be enhanced to generate enhanced prompts that may be submitted to an LLM to generate a response or responses. In such an example, the response or responses may be reviewed by one or more individuals to respond more intelligently to the series of questions of the questionnaire of the stuck pipe averter framework.

[0217] As an example, a stuck pipe averter framework may be implemented as a web application (e.g., a web app) that may pose one or more questions to anindividual, which may be deemed a questionnaire. In such an example, the questionnaire may pertain to a single well or a number of wells. In considering an example of a single well, for the individual to respond in a meaningful manner, the individual may have to resort to searching through documents (e.g., images, tables, text, etc.) associated with that single well to assure that the individual understands the history, context, etc., of the well before answering the questionnaire. In such an example, the answers to the questionnaire are consumed by the stuck pipe averter framework to generate output. In such an example, the output may depend on the quality of the answers, which, as explained, may depend on knowledge of the individual, time provided to the individual to gain knowledge, etc. In various instances where the individual’s level of knowledge is low (e.g., for lack of memory, lack of time, lack of access to documents, etc.), the output of the stuck pipe averter framework may be of greater uncertainty or less relevance. In the example of FIG. 15, the chat framework may improve operation of the stuck pipe averter framework by enhancing a query, gaining access to documents, etc., to generate output such that an individual may be able to more expeditiously respond to a questionnaire or more effectively respond to a questionnaire of the stuck pipe averter framework. In such an example, the query supplied to the chat framework may be a question of the questionnaire and / or based at least in part thereon.

[0218] As an example, in various instances, the system 1500 of FIG. 15 may be automated in that a human-in-the-loop (HITL) may not necessarily be required. In such an example, output may be a control instruction as determined by the stuck pipe averter framework based on processing of a query or queries of a questionnaire by the chat framework. In such an example, a GUI may be rendered that allows a HITL to review the control instruction prior to implementation, for example, to confirm implementation or not. In such an example, where the HITL confirms implementation, the GUI may include a graphical control actuatable by the HITL that upon actuation causes transmission of the control instruction to one or more equipment controllers.

[0219] As an example, a chat framework may be operable as an assistant to another framework that poses questions. For example, consider a chat framework that operates as an assistant to a drilling operations framework where the drilling operations framework may provide for outputting one or more control instructions responsive to receipt of answers to questions. As explained, a chat framework mayprovide for expediting answer generation and / or generation of more accurate, more relevant, etc., answers. As an example, a chat framework may provide for improving operation of another framework where answers to questions depend on a relatively large body of knowledge, which may be in the form of documents (e.g. , images, tables, text, etc.). As explained, for a well, such documents may include drilling reports, exploration reports, etc., for the well and / or one or more other wells, which may be in a common field (e.g., one or more offset wells). In general, an individual may be quite challenged to maintain knowledge of operations, events, etc., as to a well where such knowledge may be generated over a period of days, months, or years. As explained, a chat framework may provide for sorting through such knowledge to provide output, where such output may include one or more links to one or more documents relevant to the output. In such an approach, an individual may readily access a linked document to understand an underlying factual basis for the output, which may, in turn, allow the individual to assess a level of confidence in the output.

[0220] As an example, a chat framework may be operatively coupled to a drilling operations framework where the chat framework may employ reinforcement learning based at least in part on information generated by the drilling operations framework. For example, a chat framework may be provided with output from a drilling operations framework such that one or more relationships may be established, learned, etc., between queries, enhanced queries, responses, answers, output, etc. As explained, document retrieval may be performed, which may be improved via reinforcement learning. In such an example, one or more documents retrieved may be more relevant, etc., such that an individual (or machine) may operate with higher confidence, more expeditiously, etc. For example, consider an approach where rather than three documents with unknown confidence being identified, a single, best document is identified, which may result in less review time (e.g., for an individual or a machine).

[0221] As an example, a computational framework, which may be a drilling operations framework, may provide for a probabilistic decision tree as a type of output. In such an example, the output may be indicative of an issue, which may be a root cause of an issue. As explained, output may be indicative of a solution to an issue, which may be, for example, one or more control actions as to one or more drilling operations. As explained, a chat framework may expedite input generation for such acomputational framework. As explained, expediting input generation may utilize one or more machine learning models, such as, for example, one or more LLMs.

[0222] As an example, a system may be implemented using one or more guardrails. For example, consider one or more quality control assessment components, which may, for example, be implemented as part of a drilling operations framework. For example, consider an assessment component that may be operable to reject an answer to a question if the answer is deemed to be inaccurate, irrelevant, etc. (e.g., consider utilization of one or more acceptance criteria, etc.).

[0223] FIG. 16 shows example queries and responses as may be generated by a system such as, for example, the system 1500 of FIG. 15. For example, consider the scenarios below:Q1 : Any mention of cuttings volumes at surfaceR1 : Yes, there were mentions of cuttings volumes at the surface. For example, in the file "GOM_Well_A - Daily Operations 3-25-2013 05-00 - 3-26-2013 05- 00. txt", it was noted that there was a significant increase of cuttings at 1 st bottoms up returns and a noticeable decrease after 1 .5 bottoms up. In the same file, it was also observed that there was a light amount of cuttings volume at shakers during bottoms up #3.Document Reference: x1 , x5Q2: Were there torque spikes while drilling?R2: Yes, a torque spike was observed while drilling. Specifically, while back reaming out of hole at 3489 m, a torque spike to 43 kNm was observed and rotation stopped, indicating that the Bottom Hole Assembly (BHA) was stuck. Document Reference: y7Q3: Any mention of using / used LCM (lost circulation material)?R3: Yes, there were mentions of using Lost Circulation Material (LCM) in the operations on 3-23-2013 and 3-24-2013. The specific LCM used was Safe Carb 40.Document Reference: z4Q4: Any note of stuck pipe while drilling?R4: Yes, there were multiple instances of a pipe getting stuck while drilling. For example, on 9 / 5 / 2013, while pulling out of the hole at 3489m, a torque spike to 43 kNm was observed and rotation stopped, indicating that the drillpipe was stuck. Attempts to regain rotation and free the drillpipe were unsuccessful. Document Reference: x11

[0224] FIG. 17 shows example GUIs 1710 and 1720. As shown, the GUI 1710 may provide for user interaction, along with visualization as to information regarding one or more document stores, document references, etc. As an example, a user may select a document as may be referenced in a response to have the document accessed and rendered to a display for review. As shown, the GUI 1720 may provide for rendering statistics, probabilities, etc., for one or more wells.

[0225] FIG. 18 shows an example of a method 1800 that a reception block 1810 for receiving a query from a drilling operations framework; an enhancement block 1820 for enhancing the query by retrieving data from a database to form an enhanced query; a generation block 1830 for generating a response to the enhanced query using a large language model; and a process block 1840 for processing the query, based at least in part on the response, using the drilling operations framework to formulate a control action for drilling operations.

[0226] FIG. 18 also shows various computer-readable media (CRM) blocks 1811 , 1821 , 1831 , and 1841 . Such blocks may include instructions that are executable by one or more processors, which may be one or more processors of a computational framework, a system, a computer, etc. A computer-readable medium may be a computer-readable storage medium that is not a signal, not a carrier wave and that is non-transitory. For example, a computer-readable medium may be a physical memory component that may store information in a digital format.

[0227] In the example of FIG. 18, a system 1890 includes one or more information storage devices 1891 , one or more computers 1892, one or more networks 1895 and instructions 1896. As to the one or more computers 1892, each computer may include one or more processors (e.g., or processing cores) 1893 and a memory 1894 for storing the instructions 1896, for example, executable by at least one of the one or more processors. As an example, a computer may include one or more networkinterfaces (e.g., wired or wireless), one or more graphics cards, a display interface (e.g., wired or wireless), etc. The system 1890 may be specially configured to perform one or more portions of the method 1800 of FIG. 18.

[0228] As to types of machine learning models, consider one or more of a support vector machine (SVM) model, a k-nearest neighbors (KNN) model, an ensemble classifier model, a neural network (NN) model, etc. As an example, a machine learning model may be a deep learning model (e.g., deep Boltzmann machine, deep belief network, convolutional neural network, stacked auto-encoder, etc.), an ensemble model (e.g., random forest, gradient boosting machine, bootstrapped aggregation, AdaBoost, stacked generalization, gradient boosted regression tree, etc.), a neural network model (e.g., radial basis function network, perceptron, back-propagation, Hopfield network, etc.), a regularization model (e.g., ridge regression, least absolute shrinkage and selection operator, elastic net, least angle regression), a rule system model (e.g., cubist, one rule, zero rule, repeated incremental pruning to produce error reduction), a regression model (e.g., linear regression, ordinary least squares regression, stepwise regression, multivariate adaptive regression splines, locally estimated scatterplot smoothing, logistic regression, etc.), a Bayesian model (e.g., naive Bayes, average on-dependence estimators, Bayesian belief network, Gaussian naive Bayes, multinomial naive Bayes, Bayesian network), a decision tree model (e.g., classification and regression tree, iterative dichotomiser 3, C4.5, C5.0, chi-squared automatic interaction detection, decision stump, conditional decision tree, M5), a dimensionality reduction model (e.g., principal component analysis, partial least squares regression, Sammon mapping, multidimensional scaling, projection pursuit, principal component regression, partial least squares discriminant analysis, mixture discriminant analysis, quadratic discriminant analysis, regularized discriminant analysis, flexible discriminant analysis, linear discriminant analysis, etc.), an instance model (e.g., k-nearest neighbor, learning vector quantization, self-organizing map, locally weighted learning, etc.), a clustering model (e.g., k-means, k-medians, expectation maximization, hierarchical clustering, etc.), etc.

[0229] As an example, a machine model, which may be a machine learning model (ML model), may be built using a computational framework with a library, a toolbox, etc., such as, for example, those of the MATLAB framework (MathWorks, Inc.,Natick, Massachusetts). The MATLAB framework includes a toolbox that provides supervised and unsupervised machine learning algorithms, including support vector machines (SVMs), boosted and bagged decision trees, k-nearest neighbor (KNN), k- means, k-medoids, hierarchical clustering, Gaussian mixture models, and hidden Markov models. Another MATLAB framework toolbox is the Deep Learning Toolbox (DLT), which provides a framework for designing and implementing deep neural networks with algorithms, pretrained models, and apps. The DLT provides convolutional neural networks (ConvNets, CNNs) and long short-term memory (LSTM) networks to perform classification and regression on image, time-series, and text data. The DLT includes features to build network architectures such as generative adversarial networks (GANs) and Siamese networks using custom training loops, shared weights, and automatic differentiation. The DLT provides for model exchange various other frameworks.

[0230] As an example, the TENSORFLOW framework (Google LLC, Mountain View, CA) may be implemented, which is an open-source software library for dataflow programming that includes a symbolic math library, which may be implemented for machine learning applications that may include neural networks. As an example, the CAFFE framework may be implemented, which is a DL framework developed by Berkeley Al Research (BAIR) (University of California, Berkeley, California). As another example, consider the SCIKIT platform (e.g., scikit-learn), which utilizes the PYTHON programming language. As an example, a framework such as the APOLLO Al framework may be utilized (APOLLO.AI GmbH, Germany). As an example, a framework such as the PYTORCH framework may be utilized (Facebook Al Research Lab (FAIR), Facebook, Inc., Menlo Park, California).

[0231] As an example, a training method may include various actions that may operate on a dataset to train a ML model. As an example, a dataset may be split into training data and test data where test data may provide for evaluation. A method may include cross-validation of parameters and best parameters, which may be provided for model training.

[0232] The TENSORFLOW framework may run on multiple CPUs and GPUs (with optional CUDA (NVIDIA Corp., Santa Clara, California) and SYCL (The Khronos Group Inc., Beaverton, Oregon) extensions for general-purpose computing on graphics processing units (GPUs)). TENSORFLOW is available on 64-bit LINUX,MACOS (Apple Inc., Cupertino, California), WINDOWS (Microsoft Corp., Redmond, Washington), and mobile computing platforms including ANDROID (Google LLC, Mountain View, California) and IOS (Apple Inc.) operating system-based platforms.

[0233] TENSORFLOW computations may be expressed as stateful dataflow graphs; noting that the name TENSORFLOW derives from the operations that such neural networks perform on multidimensional data arrays. Such arrays may be referred to as “tensors”.

[0234] As an example, a method can include receiving a query from a drilling operations framework; enhancing the query by retrieving data from a database to form an enhanced query; generating a response to the enhanced query using a large language model; and processing the query, based at least in part on the response, using the drilling operations framework to formulate a control action for drilling operations. In such an example, the drilling operations framework may be or include a stuck pipe assessment framework. As an example, a query may be a query generated by a drilling operations framework.

[0235] As an example, a drilling operations framework may be or include a probabilistic graph based on historical data. In such an example, the probabilistic graph may be or include a belief network. As an example, a probabilistic graph may include nodes for assessment of one or more physical risks associated with drilling operations.

[0236] As an example, a control action for drilling operations may aim to reduce risk of occurrence of stuck pipe of a drillstring disposed at least in part in a borehole in a geologic environment.

[0237] As an example, a database may be or include a vector database. In such an example, a method may include retrieving data from the vector database by performing a similarity search. In such an example, a relevance metric or metrics may be generated associated with one or more search results.

[0238] As an example, a method may include retrieving data from a database that includes retrieving data from one or more documents. In such an example, a response to an enhance query may include a citation or citations to the one or more documents. As an example, a response may include a link or other data structure, object, etc., that may provide for accessing one or more documents and / or content therein (e.g., an image, a drawing, a table, text, a control instruction, etc.).

[0239] As an example, a database may include drilling reports, which may be directly and / or indirectly included. As an example, a database may include representations of drilling reports, for example, consider vector representations.

[0240] As an example, a database may include vector embeddings derived from images, tabular data, and / or textual data; noting that one or more other types of data, content, information, etc., may be provided in the form of vector embeddings (e.g., as derived from such source documentation, etc.).

[0241] As an example, data may include contextual data. As an example, contextual data may enhance a query to form an enhanced query. As an example, an LLM interface may receive a query and an enhanced query and provide different responses to each where, for example, a difference may be noted and / or otherwise highlighted. In such an example, through query enhancement, the response to the enhanced query may be expected to be more accurate and / or timely than the response to the non-enhanced query.

[0242] As an example, a method may include generating an unenhanced response to the query using the large language model and, for example, comparing the response and the unenhanced response.

[0243] As an example, a method may include enhancing a query and generating a response to the enhanced query that occur automatically in response to receiving a query.

[0244] As an example, a system may include a processor; a memory accessible by the processor; processor-executable instructions stored in the memory and executable to instruct the system to: receive a query from a drilling operations framework; enhance the query by retrieving data from a database to form an enhanced query; generate a response to the enhanced query using a large language model; and process the query, based at least in part on the response, using the drilling operations framework to formulate a control action for drilling operations.

[0245] As an example, one or more computer-readable storage media can include processor-executable instructions to instruct a computing system to: receive a query from a drilling operations framework; enhance the query by retrieving data from a database to form an enhanced query; generate a response to the enhanced query using a large language model; and process the query, based at least in part on theresponse, using the drilling operations framework to formulate a control action for drilling operations.

[0246] As an example, a method may be implemented in part using computer- readable media (CRM), for example, as a module, a block, etc. that include information such as instructions suitable for execution by one or more processors (or processor cores) to instruct a computing device or system to perform one or more actions. As an example, a single medium may be configured with instructions to allow for, at least in part, performance of various actions of a method. As an example, a computer- readable medium (CRM) may be a computer-readable storage medium (e.g., a non- transitory medium) that is not a carrier wave. As an example, a computer-program product may include instructions suitable for execution by one or more processors (or processor cores) where the instructions may be executed to implement at least a portion of a method or methods.

[0247] According to an embodiment, one or more computer-readable media may include computer-executable instructions to instruct a computing system to output information for controlling a process. For example, such instructions may provide for output to sensing process, an injection process, drilling process, an extraction process, an extrusion process, a pumping process, a heating process, etc.

[0248] In some embodiments, a method or methods may be executed by a computing system. FIG. 19 shows an example of a system 1900 that may include one or more computing systems 1901-1 , 1901-2, 1901-3 and 1901-4, which may be operatively coupled via one or more networks 1909, which may include wired and / or wireless networks.

[0249] As an example, a system may include an individual computer system or an arrangement of distributed computer systems. In the example of FIG. 19, the computer system 1901-1 may include one or more modules 1902, which may be or include processor-executable instructions, for example, executable to perform various tasks (e.g., receiving information, requesting information, processing information, simulation, outputting information, etc.).

[0250] As an example, a module may be executed independently, or in coordination with, one or more processors 1904, which is (or are) operatively coupled to one or more storage media 1906 (e.g., via wire, wirelessly, etc.). As an example, one or more of the one or more processors 1904 may be operatively coupled to atleast one of one or more network interface 1907. In such an example, the computer system 1901-1 may transmit and / or receive information, for example, via the one or more networks 1909 (e.g., consider one or more of the Internet, a private network, a cellular network, a satellite network, etc.). As shown, one or more other components 1908 may be included in the computer system 1901-1.

[0251] As an example, the computer system 1901-1 may receive from and / or transmit information to one or more other devices, which may be or include, for example, one or more of the computer systems 1901-2, etc. A device may be located in a physical location that differs from that of the computer system 1901-1. As an example, a location may be, for example, a processing facility location, a data center location (e.g., serverfarm, etc.), a rig location, a wellsite location, a downhole location, etc.

[0252] As an example, a processor may be or include a microprocessor, microcontroller, processor module or subsystem, programmable integrated circuit, programmable gate array, or another control or computing device.

[0253] As an example, the storage media 1906 may be implemented as one or more computer-readable or machine-readable storage media. As an example, storage may be distributed within and / or across multiple internal and / or external enclosures of a computing system and / or additional computing systems.

[0254] As an example, a storage medium or storage media may include one or more different forms of memory including semiconductor memory devices such as dynamic or static random access memories (DRAMs or SRAMs), erasable and programmable read-only memories (EPROMs), electrically erasable and programmable read-only memories (EEPROMs) and flash memories, magnetic disks such as fixed, floppy and removable disks, other magnetic media including tape, optical media such as compact disks (CDs) or digital video disks (DVDs), BLUERAY disks, or other types of optical storage, or other types of storage devices.

[0255] As an example, a storage medium or media may be located in a machine running machine-readable instructions, or located at a remote site from which machine-readable instructions may be downloaded over a network for execution.

[0256] As an example, various components of a system such as, for example, a computer system, may be implemented in hardware, software, or a combination ofboth hardware and software (e.g., including firmware), including one or more signal processing and / or application specific integrated circuits.

[0257] As an example, a system may include a processing apparatus that may be or include a general-purpose processors or application specific chips (e.g., or chipsets), such as ASICs, FPGAs, PLDs, or other appropriate devices.

[0258] As an example, a device may be a mobile device that includes one or more network interfaces for communication of information. For example, a mobile device may include a wireless network interface (e.g., operable via IEEE 802.11 , ETSI GSM, BLUETOOTH, satellite, etc.). As an example, a mobile device may include components such as a main processor, memory, a display, display graphics circuitry (e.g., optionally including touch and gesture circuitry), a SIM slot, audio / video circuitry, motion processing circuitry (e.g., accelerometer, gyroscope), wireless LAN circuitry, smart card circuitry, transmitter circuitry, GPS circuitry, and a battery. As an example, a mobile device may be configured as a cell phone, a tablet, etc. As an example, a method may be implemented (e.g., wholly or in part) using a mobile device. As an example, a system may include one or more mobile devices.

[0259] As an example, a system may be a distributed environment, for example, a so-called “cloud” environment where various devices, components, etc. interact for purposes of data storage, communications, computing, etc. As an example, a device or a system may include one or more components for communication of information via one or more of the Internet (e.g., where communication occurs via one or more Internet protocols), a cellular network, a satellite network, etc. As an example, a method may be implemented in a distributed environment (e.g., wholly or in part as a cloud-based service).

[0260] As an example, information may be input from a display (e.g., consider a touchscreen), output to a display or both. As an example, information may be output to a projector, a laser device, a printer, etc. such that the information may be viewed. As an example, information may be output stereographically or holographically. As to a printer, consider a 2D or a 3D printer. As an example, a 3D printer may include one or more substances that may be output to construct a 3D object. For example, data may be provided to a 3D printer to construct a 3D representation of a subterranean formation. As an example, layers may be constructed in 3D (e.g, horizons, etc.),geobodies constructed in 3D, etc. As an example, holes, fractures, etc., may be constructed in 3D (e.g., as positive structures, as negative structures, etc.).

[0261] Although only a few examples have been described in detail above, those skilled in the art will readily appreciate that many modifications are possible in the examples. Accordingly, all such modifications are intended to be included within the scope of this disclosure as defined in the following claims. In the claims, meansplus-function clauses are intended to cover the structures described herein as performing the recited function and not only structural equivalents, but also equivalent structures. Thus, although a nail and a screw may not be structural equivalents in that a nail employs a cylindrical surface to secure wooden parts together, whereas a screw employs a helical surface, in the environment of fastening wooden parts, a nail and a screw may be equivalent structures.

Claims

CLAIMSWhat is claimed is:

1. A method comprising: receiving a query from a drilling operations framework; enhancing the query by retrieving data from a database to form an enhanced query; generating a response to the enhanced query using a large language model; and processing the query, based at least in part on the response, using the drilling operations framework to formulate a control action for drilling operations.

2. The method of claim 1 , wherein the drilling operations framework comprises a stuck pipe assessment framework.

3. The method of claim 2, wherein the query comprises a query generated by the drilling operations framework.

4. The method of claim 1 , wherein the drilling operations framework comprises a probabilistic graph based on historical data.

5. The method of claim 4, wherein the probabilistic graph comprises a belief network.

6. The method of claim 4, wherein the probabilistic graph comprises nodes for assessment of physical risks associated with the drilling operations.

7. The method of claim 1 , wherein the control action for the drilling operations reduces risk of occurrence of stuck pipe of a drillstring disposed at least in part in a borehole in a geologic environment.

8. The method of claim 1 , wherein the database comprises a vector database.

9. The method of claim 8, wherein the retrieving data from the vector database comprises performing a similarity search.

10. The method of claim 1 , wherein the retrieving data from the database comprises retrieving data from one or more documents.11 . The method of claim 10, wherein the response to the enhance query comprises a citation to the one or more documents.

12. The method of claim 1 , wherein the database comprises drilling reports.

13. The method of claim 1 , wherein the database comprises vector embeddings derived from images, tabular data, and textual data.

14. The method of claim 1 , wherein the data comprises contextual data.

15. The method of claim 14, wherein the contextual data enhances the query to form the enhanced query.

16. The method of claim 1 , comprising generating an unenhanced response to the query using the large language model.

17. The method of claim 16, comprising comparing the response and the unenhanced response.

18. The method of claim 1 , wherein the enhancing and the generating occur automatically in response to the receiving.

19. A system comprising: a processor; a memory accessible by the processor; processor-executable instructions stored in the memory and executable to instruct the system to:receive a query from a drilling operations framework; enhance the query by retrieving data from a database to form an enhanced query; generate a response to the enhanced query using a large language model; and process the query, based at least in part on the response, using the drilling operations framework to formulate a control action for drilling operations.

20. One or more computer-readable storage media comprising processor-executable instructions to instruct a computing system to: receive a query from a drilling operations framework; enhance the query by retrieving data from a database to form an enhanced query; generate a response to the enhanced query using a large language model; and process the query, based at least in part on the response, using the drilling operations framework to formulate a control action for drilling operations.

Images