LIQUID LOADING IDENTIFICATION

Abstract

This disclosure relates to collecting and analyzing data associated with a well site to predict instances of liquid loading that occurs (or that will occur) in a particular well environment. Systems herein involve training a liquid loading classification model based on historical data to predict whether or not liquid loading is occurring or will occur based on real-time sensor data that is collected by a plurality of sensors at the well site. The data is preprocessed and labeled in a manner that provides an efficient training process and which enables a machine learning model, such as a recursive neural network (RNN), to accurately predict when liquid loading will occur such that a presentation may be generated and presented to an individual with the ability to prevent or otherwise mitigate the liquid loading at a time when inefficiencies and other issues can be prevented.

Description

BACKGROUND

Liquid loading is a common issue in gas wells where the velocity of gas is often insufficient to carry liquids to the surface, causing those liquids to accumulate in a wellbore. Liquid loading in these tight gas wells has troubled the industry for some time causing productivity loss and costly workovers. In conventional systems, site-specific experts are typically assigned to identify and mitigate effects of liquid loading using time and energy consuming processes. In addition, due to the unique nature of individual sites, conventional models relied on by these experts are often ineffective in addressing liquid loading at other sites. Moreover, even with domain-specific knowledge, these experts and the static models on which they rely are often unable to identify instances of liquid loading in time when mitigation can be performed and interruption of production can be avoided.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 illustrates an example environment in which drilling may occur in accordance with one or more embodiments.

FIG. 2 illustrates an example of a wellsite system in accordance with one or more embodiments.

FIG. 3 illustrates a schematic view of such an example computing or processor system, in accordance with one or more embodiments.

FIG. 4 illustrates an example computing device having a liquid loading detection and mitigation system in accordance with one or more embodiments.

FIGS. 5A-5B illustrates an example workflow showing a procedure for performing real-time liquid loading identification and mitigation in accordance with one or more embodiments.

FIGS. 6A-6C illustrate example stages of labeling liquid loading data in accordance with one or more embodiments.

FIG. 7 illustrates an example structure of a liquid loading classification model used in identifying and labeling instances of liquid loading in accordance with one or more embodiments.

FIG. 8 illustrates an example presentation generated by the liquid loading identification and mitigation system in accordance with one or more embodiments.

FIG. 9 illustrates an example series of acts for detecting and mitigating liquid loading in accordance with one or more embodiments.

DETAILED DESCRIPTION

This disclosure relates to collecting and analyzing data associated with a well site to predict instances of liquid loading that occurs (or that will occur) in a particular well environment. In one or more embodiments, a liquid loading classification model is trained based on historical data to predict whether or not liquid loading is occurring or will occur based on real-time sensor data that is collected by a plurality of sensors at the well site. As will be discussed in further detail below, the data is preprocessed and labeled in a manner that provides an efficient training process and which enables a machine learning model, such as a recursive neural network (RNN) to accurately predict when liquid loading will occur such that a presentation may be generated and presented to an individual with the ability to prevent or otherwise mitigate the liquid loading at a time when inefficiencies and other issues can be prevented.

As an illustrative example, systems, methods, and computer-readable media (or simply “systems”) described herein involve collecting historical data associated with a well site. The historical data includes information associated with instances of liquid loading corresponding to historical data samples. The systems additionally involve training a liquid loading classification model (e.g., a machine learning model, such as an RNN) based on the data samples and associated labels from the historical data. The systems further involve obtaining input data samples that include configured data (e.g., sensor data) to be provided as input to the liquid loading classification model. The systems also include applying the liquid loading classification model to the input sample data to generate an input including liquid loading labels that indicate predicted instances of liquid loading in connection with sensor data captured in connection with operation of the wellsite.

The present disclosure includes features and functionalities that implement practical applications that provide benefits and/or solve problems associated with detecting or otherwise predicting instances of liquid loading in connection with data that is captured at a wellsite. Examples of some of these applications and benefits are discussed in further detail below.

As an example, one or more embodiments described herein involves training a machine learning model that can be applied across multiple well sites. For instance, by training a model based on sensor data, the well condition, and various types of static data, such as tubing size, embodiments described herein facilitate training and implementing a liquid loading classification model that can be used across multiple well sites. This is an improvement over conventional approaches in which models and algorithms are specifically determined for individual well sites based on experts having domain-specific knowledge associated with the respective well sites. Indeed, by training a machine learning model using features and characteristics, such as well condition, sampled data, and static data for different wells (e.g., tubing size) the systems described herein provide a flexible and scalable liquid loading classification model that is capable of being implemented across different well sites.

In an effort to avoid potential inaccuracies due to the site-neutral model, one or more embodiments described herein involve providing feedback with respect to individual well sites that enables a given liquid loading classification model to be fine-tuned for an individual well site over time. Indeed, even where a liquid loading classification model is trained generally, feedback for specific sites may be considered individually in fine-tuning or otherwise retraining the liquid loading classification model to be more accurate with respect to a well site that may have unique characteristics from other well sites.

In one or more embodiments described herein, sensor data is collected and pre-processed as part of the implementation and/or training stage of the liquid loading classification model. This pre-processing or configuring of the data to be used by the liquid loading classification model facilitates a boost in accuracy related to both training and implementing the liquid loading classification model. For example, by performing various pre-processing steps, such as preliminary labels, removing outliers, and filtering noise, the systems described herein enable more efficient training of the liquid loading classification model as well as more accurate predictions of liquid loading instances by the liquid loading classification model over time. This reduces the processing expense of training the model as well as enhancing the accuracy of the model(s) itself.

As noted above, the liquid loading classification model may refer to a machine learning model that is specifically trained to generate an output including predicted instances of liquid loading with respect to a well site. In one or more embodiments, an RNN has been determined to generate more accurate predictions than other types of machine learning models. Thus, by implementing an RNN rather than other types of models, one or more embodiments described herein provide more accurate results than other conventional methods that have been used to predict or otherwise detect instances of liquid loading.

In addition, one or more embodiments described herein involve using a combination of models and algorithms in a unique way that yields more accurate and reliable predictions of liquid loading. For example, by using a combination of analytical tools and human feedback in association with the trained liquid loading classification model (e.g., a machine learning model, such as an RNN), one or more embodiments described herein yield significantly more accurate results than other conventional approaches that attempt to make brute-force predictions of liquid loading through machine learning alone or using stand-alone analytical tools. Moreover, by using analytical tools that use fewer processing resources than conventional machine learning models, the systems described herein reduce processing overhead by feeding data that has already been pre-processed and (in some instances) pre-labeled to a machine learning model that is simply confirming labels rather than attempting to determine labels from raw or unprocessed data.

Additional detail will now be provided regarding systems and methods described herein in relation to illustrative figures portraying example implementations. Wherever convenient, the same reference numbers are used in the drawings and the following description to refer to the same or similar parts. While several embodiments and features of the present disclosure are described herein, modifications, adaptations, and other implementations are possible, without departing from the spirit and scope of the present disclosure.

Although the terms “first”, “second”, etc. may be used herein to describe various elements, these terms are used to distinguish one element from another. For example, a first object or step could be termed a second object or step, and, similarly, a second object or step could be termed a first object or step, without departing from the scope of the present disclosure. The first object or step, and the second object or step, are both objects or steps respectively, but they are not to be considered the same object or step.

The terminology used in the description herein is for the purpose of describing particular embodiments and is not intended to be limiting. As used in this description and the appended claims, the singular forms “a,” “an” and “the” are intended to include the plural forms as well, unless the context clearly indicates otherwise. It will also be understood that the term “and/or” as used herein refers to and encompasses any possible combinations of one or more of the associated listed items. It will be further understood that the terms “includes,” “including,” “comprises” and/or “comprising,” when used in this specification, specify the presence of stated features, integers, steps, operations, elements, and/or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and/or groups thereof. Further, as used herein, the term “if” may be construed to mean “when” or “upon” or “in response to determining” or “in response to detecting,” depending on the context.

FIG. 1 illustrates one example of an environment 100 in which drilling may occur. The environment may include a reservoir 102 and various geological features, such as stratified layers. The geological aspects of the environment 100 may contain other features such as faults, basins, and others. The reservoir 102 may be located on land or offshore.

The environment 100 may be outfitted with sensors, detectors, actuators, etc. to be used in connection with the drilling process. FIG. 1 illustrates equipment 104 associated with a well 106 being constructed using downhole equipment 108. The downhole equipment 108 may be, for example, part of a bottom hole assembly (BHA). The BHA may be used to drill the well 106. The downhole equipment 108 may communicate information to the equipment 104 at the surface, and may receive instructions and information from the surface equipment 104 as well. The surface equipment 104 and the downhole equipment 108 may communicate using various communications techniques, such as mud-pulse telemetry, electromagnetic (EM) telemetry, or others depending on the equipment and technology in use for the drilling operation.

The surface equipment 104 may also include communication means to communicate over a network 110 to remote computing devices 112. For example, the surface equipment 104 may communicate data using a satellite network to computing devices 112 supporting a remote team monitoring and assisting in the creation of the well 106 and other wells in other locations. Depending on the communications infrastructure available at the wellsite, various communication equipment and techniques (cellular, satellite, wired Internet connection, etc.) may be used to communicate data from the surface equipment 104 to the remote computing devices 112. In some embodiments, the surface equipment 104 sends data from measurements taken at the surface and measurements taken downhole by the downhole equipment 108 to the remote computing devices 112.

During the well construction process, a variety of operations (such as cementing, wireline evaluation, testing, etc.) may also be conducted. In such embodiments, the data collected by tools and sensors and used for reasons such as reservoir characterization may also be collected and transmitted by the surface equipment 104.

In FIG. 1, the well 106 includes a substantially horizontal portion (e.g., lateral portion) that may intersect with one or more fractures. For example, a well in a shale formation may pass through natural fractures, artificial fractures (e.g., hydraulic fractures), or a combination thereof. Such a well may be constructed using directional drilling techniques as described herein. However, these same techniques may be used in connection with other types of directional wells (such as slant wells, S-shaped wells, deep inclined wells, and others) and are not limited to horizontal wells.

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 can include a mud tank 201 for holding mud and other material (e.g., where mud can 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, a derrick 214 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 preventors (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.

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 one or more directional drilling techniques, equipment, etc.

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).

The wellsite system 200 can provide for operation of the drillstring 225 and other operations. As shown, the wellsite system 200 includes the traveling block 211 and the derrick 214 positioned over the borehole 232. As mentioned, the wellsite system 200 can include the rotary table 220 where the drillstring 225 passes through an opening in the rotary table 220.

As shown in the example of FIG. 2, the wellsite system 200 can 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 can 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 can pass through the kelly drive bushing 219, which can be driven by the rotary table 220. As an example, the rotary table 220 can include a master bushing that operatively couples to the kelly drive bushing 219 such that rotation of the rotary table 220 can turn the kelly drive bushing 219 and hence the kelly 218. The kelly drive bushing 219 can 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 can freely move up and down inside the kelly drive bushing 219.

As to a top drive example, the top drive 240 can provide functions performed by a kelly and a rotary table. The top drive 240 can turn the drillstring 225. As an example, the top drive 240 can 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 can 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.

In the example of FIG. 2, the mud tank 201 can hold mud, which can 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.).

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., or other 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 can 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 can 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.).

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.

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 can 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.

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.

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.).

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 can 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.

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.

The assembly 250 of the illustrated example includes a logging-while-drilling (LWD) module 254, a measurement-while-drilling (MWD) module 256, an optional module 258, a rotary-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 can include a plurality of tools.

As to a 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 can commence with a vertical portion and then deviate from vertical such 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.

One approach to directional drilling involves a mud motor; however, a mud motor can 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 can 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.

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 can 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.

As an example, a PDM mud motor can operate in a so-called sliding mode, when the drillstring is not rotated from the surface. In such an example, a bit RPM can be determined or estimated based on the RPM of the mud motor.

A RSS can drill directionally where there is continuous rotation from surface equipment, which can alleviate the sliding of a steerable motor (e.g., a PDM). A RSS may be deployed when drilling directionally (e.g., deviated, horizontal, or extended-reach wells). A RSS can aim to minimize interaction with a borehole wall, which can help to preserve borehole quality. A RSS can 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.

The LWD module 254 may be housed in a suitable type of drill collar and can 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 can be employed, for example, as represented by the module 256 of the drillstring assembly 250. Where the position of an LWD module is mentioned, as an example, it may refer to a module at the position of the LWD module 254, the module 256, etc. An LWD module can 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.

The MWD module 256 may be housed in a suitable type of drill collar and can 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 can 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.

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.

As an example, a drilling operation can 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.

As an example, a directional well can 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.

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

As an example, a system may be a steerable system and include equipment to perform method such as geosteering. As mentioned, a steerable system can be or include an RSS. As an example, a steerable system can include a PDM or a turbine on a lower part of a drillstring which, just above a drill bit, a bent sub can 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 can 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.).

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, can allow for implementing a geosteering method. Such a method can include navigating a subsurface environment, for example, to follow a desired route to reach a desired target or targets.

As an example, a drillstring can 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 optional equipment for measuring and/or responding to one or more of inclination, resistivity and gamma ray related phenomena.

As an example, geosteering can 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 keep the 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.

Referring again to FIG. 2, the wellsite system 200 can 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 sensors 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).

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

As an example, the system 200 can include one or more sensors 266 that can 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 can be operatively coupled to portions of the standpipe 208 through which mud flows. As an example, a downhole tool can generate pulses that can 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 can include associated circuitry such as, for example, encoding circuitry that can 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 can include a transmitter that can generate signals that can be transmitted downhole via mud (e.g., drilling fluid) as a transmission medium.

As an example, one or more portions of a drillstring may become stuck. The term stuck can 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 amount of 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.

As to the term “stuck pipe”, this can refer to a portion of a drillstring that cannot be rotated or moved axially. As an example, a condition referred to as “differential sticking” can 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 can have time and financial cost.

As an example, a sticking force can 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 can be just as effective in sticking pipe as can a high differential pressure applied over a small area.

As an example, a condition referred to as “mechanical sticking” can be a condition where limiting or preventing motion of the drillstring by a mechanism other than differential pressure sticking occurs. Mechanical sticking can 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.

FIG. 3 illustrates a schematic view of such a computing or processor system 300, according to an embodiment. The processor system 300 may include one or more processors 302 of varying core configurations (including multiple cores) and clock frequencies. The one or more processors 302 may be operable to execute instructions, apply logic, etc. It will be appreciated that these functions may be provided by multiple processors or multiple cores on a single chip operating in parallel and/or communicably linked together. In at least one embodiment, the one or more processors 302 may be or include one or more GPUs.

The processor system 300 may also include a memory system, which may be or include one or more memory devices 304 and/or computer-readable media of varying physical dimensions, accessibility, storage capacities, etc. such as flash drives, hard drives, disks, random access memory, etc., for storing data, such as images, files, and program instructions for execution by the processor 302. In an embodiment, the computer-readable media may store instructions that, when executed by the processor 302, are configured to cause the processor system 300 to perform operations. For example, execution of such instructions may cause the processor system 300 to implement one or more portions and/or embodiments of the method(s) described above.

The processor system 300 may also include one or more network interfaces 306. The network interfaces 306 may include any hardware, applications, and/or other software. Accordingly, the network interfaces 306 may include Ethernet adapters, wireless transceivers, PCI interfaces, and/or serial network components, for communicating over wired or wireless media using protocols, such as Ethernet, wireless Ethernet, etc.

As an example, the processor system 300 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 one or more IEEE 802.11 protocols, 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.

The processor system 300 may further include one or more peripheral interfaces 308 for communication with a display, projector, keyboards, mice, touchpads, sensors, other types of input and/or output peripherals, and/or the like. In some implementations, the components of processor system 300 need not be enclosed within a single enclosure or even located in close proximity to one another, but in other implementations, the components and/or others may be provided in a single enclosure. 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 method may be implemented in a distributed environment (e.g., wholly or in part as a cloud-based service).

As an example, information may be input from a display (e.g., 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 can 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.).

The memory device 304 may be physically or logically arranged or configured to store data on one or more storage devices 310. The storage device 310 may include one or more file systems or databases in any suitable format. The storage device 310 may also include one or more software programs 312, which may contain interpretable or executable instructions for performing one or more of the disclosed processes. When requested by the processor 302, one or more of the software programs 312, or a portion thereof, may be loaded from the storage devices 310 to the memory devices 304 for execution by the processor 302.

Those skilled in the art will appreciate that the above-described componentry is merely one example of a hardware configuration, as the processor system 300 may include any type of hardware components, including any accompanying firmware or software, for performing the disclosed implementations. The processor system 300 may also be implemented in part or in whole by electronic circuit components or processors, such as application-specific integrated circuits (ASICs) or field-programmable gate arrays (FPGAs).

The processor system 300 may be configured to receive a variety of data associated with a particular wellsite. Examples include drilling data, well data, production data, pressure data, etc. The received or otherwise obtained data may be collected using one or more sensors associated with surface equipment or with downhole equipment. The various data may include information such as data relating to the position of the BHA (such as survey data or continuous position data), drilling parameters (such as weight on bit (WOB), rate of penetration (ROP), torque, or others), production metrics, text information entered by individuals working at the wellsite, or other data collected during the construction of or operation of the well.

In one embodiment, the processor system 300 is part of a rig control system (RCS) for the rig. In another embodiment, the processor system 300 is a separately installed computing unit including a display that is installed at the rig site and receives data from the RCS. In such an embodiment, the software on the processor system 300 may be installed on the computing unit, brought to the wellsite, and installed and communicatively connected to the rig control system in preparation for constructing the well or a portion thereof.

In another embodiment, the processor system 300 may be at a location remote from the wellsite and receives the various data over a communications medium using a protocol such as well-site information transfer specification or standard (WITS) and markup language (WITSML). In such an embodiment, the software on the processor system 300 may be a web-native application that is accessed by users using a web browser. In such an embodiment, the processor system 300 may be remote from the wellsite where the well is being constructed, and the user may be at the wellsite or at a location remote from the wellsite.

While the discussion above references drilling, the approaches discussed herein may be applied to other wellsite operations. These include, but are not limited to, wireline operations, production operations (such as artificial lift operations), and others. Similarly, the various data may be data from other wellsite operations.

As noted above, liquid loading in tight gas wells has troubled the industry for a long time. These events can result in productivity loss and costly workovers. Dedicated people are frequently given the dedicated task of identifying and mitigating liquid loading effects, which is very time and energy consuming. Discussed herein is a deep learning approach to identify liquid loading in real time. The approach helps operators manage liquid loading properly, and it can relieve people from tedious work of checking the production profile all the time.

FIG. 4 illustrates a block diagram showing an example computing device 400 on which a liquid loading detection and mitigation system 402 (or simply “liquid loading system 402”) may be implemented in accordance with one or more embodiments. The computing device may refer to a client device, server device, or any other type of computing device on which a liquid loading system 402 may be implemented. For example, the computing device 400 may refer to a mobile phone, smartphone, tablet, desktop, or non-mobile server device. In one or more embodiments, the computing device 400 is implemented on a cloud computing system. In one or more embodiments, the computing device 400 is a singular device in which each of the components of the liquid loading system 402 are implemented thereon. In other examples, one or more components of the liquid loading system 402 may be implemented across different computing devices. For example, in one or more embodiments, training of a liquid loading classification model may be performed using a first one or more devices while implementation of the liquid loading classification model is performed using a second one or more devices.

As shown in FIG. 4, the liquid loading system 402 includes a plurality of components, which will be discussed in further detail in connection with various examples. As shown in FIG. 4, the liquid loading system 402 includes a sample collection manager 404, and a data configuration manager 406 including one or more data processing tools 407. As further shown, the liquid loading system 402 includes a model manager 408 having an instance of the liquid loading classification model 410 thereon. The liquid loading system 402 additionally includes a presentation manager 412 and a data storage 414.

It will be understood that while FIG. 4 illustrates an example in which each of the illustrated components 404-414 are implemented in whole on the computing device(s) 400, other implementations may include one or more components (or sub-components) implemented across different devices of the environment in which the computing device 400 is implemented. As a non-limiting example, the sample collection manager 404 and data configuration manager 406 may be implemented on a computing device while the model manager 408 is implemented on a server device of a cloud computing system. As another example, the model manager 408 may be implemented across multiple different devices with a training feature of the model manager 408 being implemented on a first computing device (e.g., on a cloud computing system) while implementation of the model manager 408 or a trained instance of the liquid loading classification model 410 is implemented on a second computing device (e.g., a desktop or edge server near a wellsite). Thus, while one or more embodiments described herein refer specifically to a liquid loading system 402 implemented on a single device or system of devices, features and functionalities of the liquid loading system 402 may similarly apply to other device environments.

As mentioned above, the liquid loading system 402 includes a sample collection manager 404. The sample collection manager 404 may receive, collect, or otherwise obtain access to data that is captured by any number of sensors. In one or more embodiments, the sample collection manager 404 collects historical data associated with a wellsite by capturing or accessing data that is locally captured by a plurality of sensors. In one or more embodiments, the sample collection manager 404 includes the sensor(s) tasked with capturing sensor data.

By way of example and not limitation, the sample collection manager 404 may collect sample data from a variety of sensors that provide data useful in detecting or otherwise monitoring instances of liquid loading. For example, in one or more embodiments, the sensors include tubing head pressure (THP) sensors tasked with capturing pressure measurements at a top of a tubing string. In some instances, this includes capturing temperature and flow rate of fluids and/or gasses at a location of the THP sensor(s). As another example, in one or more embodiments, the sensors include casing head pressure (CHP) sensors tasked with capturing pressure, temperature, and/or flow rate measurements at a top of a casing string. Additional example sensors include temperature sensors tasked with detecting or otherwise capturing temperature metrics at various locations within the wellsite. In one or more embodiments, the sensors include one or more gas rate sensors or flow meters capable of capturing flow rates of gasses and/or fluids at different locations within the wellsite.

The sample collection manager 404 may collect the sensor data from the respective sensors in a variety of ways. In one or more embodiments, the sensors are electrically coupled to the sample collection manager 404 via one or more wired connections. Alternatively, in one or more embodiments, the sensors are communicatively coupled to the sample collection manager 404 via one or more wireless connections (e.g., a wireless network). In one or more embodiments, the sensors capture and locally store data for later retrieval from the respective sensors by the sample collection manager 404 at a later time.

As will be discussed in further detail below, the sample collection manager 404 may be tasked with collecting sensor data in both the training and implementation stages of a liquid loading classification model 410. For example, as will be discussed in further detail below in connection with FIG. 5A, the sample collection manager 404 may collect historical data including sensor data captured over a previous period of time prior to predicting liquid loading instances in real-time. In addition, and as will be discussed in further detail below in connection with FIG. 5B, the sample collection manager 404 may collect real-time and/or current sensor data captured over a predetermined period of time (e.g., one hour) regarding data on which the liquid loading classification model 410 is to be applied to predict or otherwise determine instances of liquid loading associated with the captured sensor data.

As mentioned above, and as shown in FIG. 4, the liquid loading system 402 includes a data configuration manager 406. As noted above, the sensor data collected from the plurality of sensors may be processed (e.g., pre-processed, pre-labeled) or otherwise configured prior to feeding the data as an input to a model. While one or more embodiments described here refer specifically to configuring the data prior to feeding the data as an input to the liquid loading classification model 410 that has been previously trained, this configuration of the sensor data may be applicable to generating training data for the liquid loading classification model 410 as well as generating input data for the liquid loading classification model 410.

The data configuration manager 406 may obtain sample data. In one or more embodiments, obtaining the data includes configuring the sensor data in a variety of ways and using a variety of tools. For example, as shown in FIG. 4, the data configuration manager 406 may make use of a variety of data processing tools 407. In one or more embodiments, the data processing tools 407 refer to various analytical tools that may be applied to the sensor data to prepare the data as an input to the liquid loading classification model 410.

By way of example, one or more of the data processing tools 407 refers to an outlier removal tool. For instance, one data processing tool may refer to a software tool that removes outlier data samples from the sensor data based on predicted errors of the sensor capture based on distances or differences between values of samples captured by the respective sensors. This may involve removing outliers having at least a threshold difference from a mean value of other samples of similar types of sensor data (e.g., data captured by the same sensor or same type of sensor). The outlier removal tool may utilize any number of statistical methods or formulas in determining specific criteria for removing outliers. In one or more embodiments, the outlier removal tool relies on operator-configured parameters (e.g., configurable thresholds) in determining which data samples should be removed from the sensor data that is used to train the model(s) and/or provided as an input to the model(s).

As another example, one or more of the data processing tools 407 refers to a noise removal tool. For instance, one data processing tool may refer to a software tool that smooths or otherwise normalizes data to remove noise or fluctuations within the sensor data. Indeed, because the flow rate and/or pressure metrics may spike at certain instances, the data configuration manager 406 may smooth some of the sampled data by reducing or increasing certain values over time to create smooth or otherwise normalized data to prevent certain inaccuracies from being introduced to the liquid loading classification model 410. This provides a more efficient training process as well as more accurate outputs when providing the input data to the liquid loading classification model 410. Similar to the outlier removal tool, the noise removal tool may include parameters that are configurable and determine to what degree the data is smoothed (or not smoothed) when processing the data in preparation for training and/or applying the liquid loading classification model 410.

As another example, one or more of the data processing tools 407 refers to a data imputation tool. In one or more embodiments, the data may include periods of time in which sensor data is not captured, such as where a connection between a sensor and the sample collection manager 404 goes down or where a particular sensor simply does not collect any readings for some reason or another. In any example where sensor data is missing, a data imputation processing tool 407 may be used to generate predicted data based on observed trends of sensor data that is captured during a period around (e.g., before and/or after) when the sensor data is missing. Indeed, the data configuration manager 406 may use any number of known analytical tools or statistical methods to fill in or otherwise impute this missing data in an effort to more efficiently train and/or boost accuracy of the liquid loading classification model 410.

It will be appreciated that one or more of the data processing tools 407 may be used in combination with other data processing tools 407 to generate the configured data. For example, in one or more embodiments, the data configuration manager 406 may rely on pre-processing performed by one or more of the data processing tools 407 (e.g., an outlier removal tool, noise removal tool, and/or data imputation tool) to apply one or more additional processing tools 407 to the data samples. For instance, in one or more embodiments, the data configuration manager 406 applies a fast-labeling tool (or fast labeling model) that is specifically configured or trained to apply preliminary labels to the pre-processed samples.

As an example, in one or more embodiments, the data configuration manager 406 uses fast-labeling algorithm(s) that consider combinations of sensor measurements (e.g., pressure differences, temperature, flow rates) to calculate predicted labels for the various samples. Unlike the liquid loading classification model 410, these algorithms may refer to static or predefined algorithms that consider snapshots or limited ranges of measured values to determine predicted labels that are pre-applied to the various samples as part of a training process. Indeed, it will be appreciated that the analytical models used at this stage will likely be less robust and not a machine learning model as will likely be the case with the liquid loading classification model 410.

In one or more embodiments, the data configuration manager 406 receives additional input confirming or negating certain labels or values that have been added as part of the configuration of the sensor data prior to training the liquid loading classification model 410. For example, in one or more embodiments, the fast-labeling is reviewed by one or more experts or administrators that can confirm whether the particular labeling is accurate or not. While not necessary for all implementations, this additional human element can provide an improvement to the accuracy of the model as well as improve the efficiency with which the liquid loading classification model is trained.

As mentioned above, and as further shown in FIG. 4, the liquid loading system 402 includes a model manager 408. The model manager 408 is tasked with training and/or implementing the liquid loading classification model 410. Additional detail in connection with training and implementing the liquid loading classification model 410 will be discussed below in connection with FIGS. 5A-5B.

It will be appreciated that the liquid loading classification model 410 may refer to a variety of models and/or algorithms that are used in predicting instances of liquid loading within a production environment. In one or more embodiments, the liquid loading classification model 410 refers to a machine learning model that is trained based on training data that is collected and configured as discussed in various examples above. In one or more embodiments, the liquid loading classification model 410 refers specifically to a neural network, such as a RNN. Other example implementations include random forest models, decision tree models, and other types of machine learning models. As mentioned above, the RNN model provides enhanced accuracy over many of the existing models, in part due to the time-series nature and structure of the model itself.

In one or more embodiments described herein, the liquid loading classification model 410 is trained based on a combination of sensor data as well as one or more formulas based on the sensor data. In one or more examples, the liquid loading classification model 410 considers a severity of liquid loading in making a prediction of a particular instance of liquid loading. In one or more embodiments, the liquid loading classification model 410 determines a predicted severity of liquid loading at a given point in time based on a severity metric.

The severity metric may be determined based on a combination of metrics involving a combination of various sensor outputs. For instance, in one or more embodiments, the severity metric is determined based on an average of a gas rate metric and a pressure metric. The gas rate metric may refer to a maximum observed gas rate over a predetermined period of time (e.g., an hour). The pressure metric may be determined based on a difference between a case heading pressure (CHP) and a tubing head pressure (THP) observed by the plurality of sensors (e.g., over a period of time). In one or more embodiments, the liquid loading classification model 410 determines a severity metric based on the following formula:

${\mathrm{Rate}}_{\mathrm{indicator}}=\max \left(\frac{\mathrm{mean}\left({\mathrm{rate}}_{\mathrm{prehour}}\right)-\mathrm{rate}}{\mathrm{rate}}\right),0$ ${\mathrm{Press}}_{\mathrm{indicator}}=\frac{\mathrm{CHP}-\mathrm{THP}}{\mathrm{CHP}}$ $\mathrm{Severity}=\mathrm{AVG}\left({\mathrm{Rate}}_{\mathrm{indicator}},{\mathrm{Press}}_{\mathrm{indicator}}\right)$

Where “rate” refers to a particular gas flow rate, “press” refers to a particular pressure metric, and “severity” refers to the indicated average as a function of rate and pressure over the relevant period of time. Other formulas may be used that consider certain thresholds and metrics. In one or more embodiments, the severity metric is specifically used (e.g., in combination with other metrics) to determine an output prediction of liquid loading with respect to a particular snapshot of sensor data and/or an output prediction of liquid loading with respect to a particular range of time.

By implementing the severity metric, the liquid loading system 402 facilitates early detection of liquid loading, often at a point when mitigation can be performed without significant production losses. Indeed, as will be discussed below in connection with additional examples, the liquid loading system 402 predicts instances of liquid loading prior to halt of production due to more severe instances of liquid loading. Indeed, the severity metric provides the ability of the liquid loading system 402 to effectively predict liquid loading at an earlier stage of production than other conventional methods that rely wholly on fast-labeling, human expertise, or even machine learning-based prediction models alone.

As discussed above, and as further shown in FIG. 4, the liquid loading system 402 further includes a presentation manager 412. The presentation manager 412 is tasked with generating a presentation related to liquid loading predictions based on outputs generated by the liquid loading classification model 410. As will be discussed below, the presentation manager 412 may generate a presentation including information to be displayed via a graphical user interface (GUI) based on outputs of the liquid loading classification model 410. Examples of the data that may be displayed will be discussed below in connection with additional figures.

As further shown in FIG. 4, the liquid loading system 402 includes a data storage 414. The data storage 414 may include any information or data relied on by any of the components of the liquid loading system 402 to perform any of the acts discussed herein. For example, the data storage 414 may include model data, sensor data, data processing information, or other information that the respective components 404-412 of the liquid loading system 402 use in performing the acts described herein. As shown in FIG. 4, the data storage 414 may include information stored on the computing device(s) 400. Alternatively, the data storage 414 may refer to data stored on another device that is accessible to one or more of the components 404-412.

FIG. 5A illustrates an example workflow 500A showing an example method of training a liquid loading classification model 410 in accordance with one or more embodiments. As shown in FIG. 5A, different components of the liquid loading system 402 discussed above are implemented to perform similar acts as discussed above in connection with FIG. 4.

As shown in FIG. 5A, a plurality of sensors 504 are implemented at a well site 502. As noted above, these sensors may refer to a variety of sensors capable of detecting or otherwise capturing data associated with flow rate, pressure, temperature, etc. As shown in FIG. 5A, the sample collection manager 404 receives historical data 506 from the plurality of sensors 504. As discussed above, the sample collection manager 404 can receive this historical data 506 in a variety of ways including, by way of example, a wired connection, wireless network, or other communication medium through which the sample collection manager 404 and sensors 504 are communicatively connected.

As further shown in FIG. 5A, the sample collection manager 404 may provide the sensor data 508 to a data configuration manager 406. The data configuration manager 406 may process (e.g., pre-process, fast-label) the sensor data in a number of ways. For example, the data configuration manager 406 can remove outliers, filter or smooth the data by removing noise, and impute missing data. In one or more embodiments, the data configuration manager 406 additionally implements a fast-labeling technique in which data samples are preliminarily labeled based on one or more algorithms or analytical tools that are applied to the sensor data.

In short, the data configuration manager 406 may perform any number of operations on the sensor data to generate processed data 510 that can be used to train the liquid loading classification model 410. As shown in FIG. 5A, the liquid loading classification model 410 may be trained based on a combination of the generated processed data 510 (e.g., based on historical data collective over a previous period of time) as well as based on additional training data 511 that is provided to the liquid loading classification model 410. The additional training data may refer to manual data, such as confirmation or corrections of the various labels (e.g., pre-labels). The additional training data may also refer to training parameters such as thresholds and/or metrics of accuracy or aggressiveness with which the liquid loading classification model 410 is to be trained. Once trained, the liquid loading classification model 410 may be used on current data originating from the wellsite.

For example, FIG. 5B illustrates an example workflow 500B illustrating an example implementation of a trained liquid loading classification model 410 in accordance with one or more embodiments. Similar to FIG. 5A, the plurality of sensors 504 are shown as implemented at the well site 502. While FIG. 5B illustrates an example in which the well site 502 and the plurality of sensors 504 are the same as the wellsite and plurality of sensors discussed in connection with FIG. 5A, other implementations may involve implementing the liquid loading classification model 410 to predict instances of liquid loading at a different wellsite or in connection with different sensors.

As shown in FIG. 5B, the sample collection manager 404 receives or otherwise obtains real-time sensor data 512 from the plurality of sensors 504. In one or more embodiments, the sample collection manager 404 collects and stores a certain quantity of the sensor data 512 for a predetermined period of time (e.g., one hour, one day, multiple days).

As shown in FIG. 5B, the data configuration manager 406 obtains the sensor data 514 from the sample collection manager 404. Similar to one or more embodiments described herein, the data configuration manager 406 processes the sensor data 514 to generate processed data 516 to provide as an input to the liquid loading classification model 410. As noted above, the liquid loading classification model 410 may refer to a RNN that is trained to generate an output including a prediction of liquid loading instances.

As shown in FIG. 5B, the liquid loading classification model 410 generates an output including labels and/or other information indicating predicted instances of liquid loading in the captured sensor data. In one or more embodiments, the liquid loading classification model 410 provides the output to a presentation manager for further processing and/or for generating a presentation of the predictions generated by the liquid loading classification model 410.

As shown in FIG. 5B, a presentation device 520 may generate and display a presentation showing the predictions of liquid loading generated by the liquid loading classification model 410. In this example, the presentation indicates tracked pressure and/or temperature readings overtime as well as indications of whether the readings at the corresponding timestamps are associated with predicted instances of liquid loading.

In one or more embodiments, an individual or expert may review the predicted instances of liquid loading and provide confirmation of whether the predictions are correct or not. For example, after a period of time and/or after addressing the liquid loading, an individual (e.g., user of the presentation device 520) may provide a confirmation with respect to the identified predictions of liquid loading about whether the predictions were correct or not. As shown in FIG. 5B, the presentation device 520 may provide feedback data 522 to the liquid loading classification model 410. In one or more embodiments, the liquid loading classification model 410 uses this feedback data 522 to fine-tune one or more layers, parameters, or algorithms used by the liquid loading classification model 410 to generate the predictions of liquid loading. In one or more embodiments, the presentation device 520 presents the display via a web application interface. Alternatively, in one or more embodiments, the presentation device 520 presents the display via a stand-alone or dedicated application associated with displaying/presenting the liquid loading results.

By providing the feedback data 522, the liquid loading classification model 410 realizes a number of benefits. For example, by considering feedback data 522, the liquid loading classification model 410 improves in accuracy over time as additional training data is provided to the liquid loading classification model 410.

In addition, where a trained liquid loading classification model 410 may be deployed to different computer systems to be used in connection with multiple wellsites, the liquid loading classification model 410 may be refined differently at different sites based on different behavior of the respective wellsites. In this way, customizations of the liquid loading classification model 410 may be realized at different wellsites over time to produce liquid loading classification models that are uniquely trained based on wellsites that have slightly different (or substantially different) behaviors with respect to liquid loading.

FIGS. 6A-6C illustrate examples views showing different stages in which the sensor data can be processed. For example, FIG. 6A illustrates an example involving fast labeling of the data. In particular, FIG. 6A illustrates example results of preliminary labeling generated by analytical algorithms. As seen in FIG. 6A, the labeling by analytic methods may not be continuous. The result of the initial labeling may result in large numbers of points not being identified, and further processing may be required to process labels, correct mislabeled parts, and remove isolated labels. FIG. 6B illustrates labels after post processing steps such as those listed above. FIG. 6C illustrates one embodiment of the results after quality checking and revision. In certain embodiments, human users may quality check the results and correct inappropriate labels. The labeling and identification may not need to be 100% accurate; a small number of mis-labeling is unlikely to negatively impact the ultimate prediction results.

As discussed above, systems and methods described herein may involve the use of machine learning techniques. Candidate machine learning techniques include SVM, KNC, Logistic regression, random forest, and RNN. The RNN may be an artificial neural network that is adapted to work for time series data and/or sequential data. The RNN may be configured to store and/or leverage historical information to generate predictions based on past activity. The timeframe may be past information from hours, days, or other range of time. FIG. 7 illustrates one sample architecture in accordance with one or more embodiments.

FIG. 8 illustrates an example implementation showing one possible visualization. The visualization to the left in FIG. 8 shows labels generated by the fast-labeling technique (e.g., without the machine learning model), and those to the right show the predictions using the RNN model (e.g., in combination with the fast-labeling and other tools for processing the sensor data.

Turning now to FIG. 9, this figure illustrates example flowcharts including series of acts for predicting instances of liquid loading using a machine learning model trained and implemented in accordance with one or more embodiments described herein. While FIG. 9 illustrates acts according to one or more embodiments, alternative embodiments may omit, add to, reorder, and/or modify any of the acts shown in FIG. 9. The acts of FIG. 9 can be performed as part of a method. Alternatively, a non-transitory computer-readable medium can include instructions that, when executed by one or more processors, cause a computing device to perform the acts of FIG. 9. In still further embodiments, a system can perform the acts of FIG. 9.

As shown in FIG. 9, the series of acts 900 includes an act 910 of collecting historical data including historical data samples captured by a plurality of sensors and associated labels indicating instances of liquid loading. For example, in one or more implementations, the act 910 includes collecting historical data associated with a well site, the historical data including historical data samples captured by a plurality of sensors and associated labels indicating instances of liquid loading corresponding to the historical data samples.

As further shown, the series of acts 900 includes an act 920 of training a liquid loading classification model based on the data samples and associated labels. In one or more embodiments, the act 920 involves training a liquid loading classification model based on the data samples and associated labels from the historical data.

As further shown, the series of acts 900 includes an act 930 of obtaining input sample data by configuring sensor data captured by the plurality of sensors to be provided as input to the liquid loading classification model. In one or more embodiments, the act 930 includes obtaining input sample data, wherein obtaining the input sample data includes configuring sensor data captured by the plurality of sensors to be provided as input to the liquid loading classification model.

As further shown, the series of acts 900 includes an act 940 of applying the liquid loading classification model to the input sample data to generate an output including liquid loading labels indicating predicated instances of liquid loading in connection with the sensor data. In one or more embodiments, the act 940 involves applying the liquid loading classification model to the input sample data to generate an output including liquid loading labels indicating predicted instances of liquid loading in connection with the sensor data captured by the plurality of sensors and associated timestamps of the sensor data.

In one or more embodiments, the liquid loading classification model is a machine learning model trained to predict instances of liquid loading based on series of given data samples captured by a given plurality of sensors associated with a given well site. In one or more embodiments, the machine learning model is a recursive neural network (RNN) model.

In one or more embodiments, the plurality of sensors includes one or more tubing head pressure (THP) sensors and one or more casing head pressure (CHP) sensors. In one or more embodiments, the plurality of sensors further includes one or more gas rate sensors. In one or more embodiments, the plurality of sensors further includes one or more temperature sensors.

In one or more embodiments, configuring the sensor data captured by the plurality of sensors includes pre-processing the sensor data. This may include removing outliers from the sensor data based on distances between values of samples of the sensor data and a mean value of other samples from the sensor data. This may further include removing noise from the sensor data by applying a smoothing algorithm to reduce data fluctuations within the sensor data. In one or more embodiments, configuring the sensor data captured by the plurality of sensors further includes applying a fast-labeling model to the pre-processed sensor data to generate preliminary labels for the sensor data prior to applying the liquid loading classification model to the sampled input data.

In one or more embodiments, the liquid loading classification model is trained to predict instances of liquid loading based at least in part on a liquid loading severity metric, the liquid loading severity metric being determined based on an average of a gas rate metric and a pressure metric. In one or more embodiments, the gas rate metric is based on a maximum observed gas rate over a predetermined period of time, and wherein the pressure metric is based on a difference between a case heading pressure (CHP) and a tubing head pressure (THP) observed by the plurality of sensors.

In one or more embodiments, the series of acts 900 includes causing a computing device to generate and display a presentation including a visualization of any predicted instances of liquid loading based on the output of the liquid loading classification model.

As noted above, liquid loading is a common issue and can pose a significant obstacle for gas production. Compared to traditional solutions, the approach described above can leverage expert knowledge and AI technology to automate aspects of liquid loading detection resulting in benefits for the client.

The embodiments disclosed in this disclosure are to help explain the concepts described herein. This description is not exhaustive and does not limit the claims to the precise embodiments disclosed. Modifications and variations from the exact embodiments in this disclosure may still be within the scope of the claims.

Likewise, the steps described need not be performed in the same sequence discussed or with the same degree of separation. Various steps may be omitted, repeated, combined, or divided, as appropriate. Accordingly, the present disclosure is not limited to the above-described embodiments, but instead is defined by the appended claims in light of their full scope of equivalents. In the above description and in the below claims, unless specified otherwise, the term “execute” and its variants are to be interpreted as pertaining to any operation of program code or instructions on a device, whether compiled, interpreted, or run using other techniques.

The claims that follow do not invoke section 112 (f) unless the phrase “means for” is expressly used together with an associated function.

Claims

1. A method, comprising: collecting historical data associated with a well site, the historical data including historical data samples captured by a plurality of sensors and associated labels indicating instances of liquid loading corresponding to the historical data samples;training a liquid loading classification model based on the data samples and associated labels from the historical data;obtaining input sample data, wherein obtaining the input sample data includes configuring sensor data captured by the plurality of sensors to be provided as input to the liquid loading classification model;applying the liquid loading classification model to the input sample data to generate an output including liquid loading labels indicating predicted instances of liquid loading in connection with the sensor data captured by the plurality of sensors and associated timestamps of the sensor data.

2. The method of claim 1, wherein the liquid loading classification model is a machine learning model trained to predict instances of liquid loading based on series of given data samples captured by a given plurality of sensors associated with a given well site.

3. The method of claim 2, wherein the machine learning model is a recursive neural network (RNN) model.

4. The method of claim 1, wherein the plurality of sensors includes one or more tubing head pressure (THP) sensors and one or more casing head pressure (CHP) sensors.

5. The method of claim 4, wherein the plurality of sensors further includes one or more gas rate sensors.

6. The method of claim 4, wherein the plurality of sensors further includes one or more temperature sensors.

7. The method of claim 1, wherein configuring the sensor data captured by the plurality of sensors includes pre-processing the sensor data by performing one or more of: removing outliers from the sensor data based on distances between values of samples of the sensor data and a mean value of other samples from the sensor data;removing noise from the sensor data by applying a smoothing algorithm to reduce data fluctuations within the sensor data.

8. The method of claim 7, wherein configuring the sensor data captured by the plurality of sensors further includes applying a fast-labeling model to the pre-processed sensor data to generate preliminary labels for the sensor data prior to applying the liquid loading classification model to the sampled input data.

9. The method of claim 1, wherein the liquid loading classification model is trained to predict instances of liquid loading based at least in part on a liquid loading severity metric, the liquid loading severity metric being determined based on an average of a gas rate metric and a pressure metric.

10. The method of claim 9, wherein the gas rate metric is based on a maximum observed gas rate over a predetermined period of time, and wherein the pressure metric is based on a difference between a case heading pressure (CHP) and a tubing head pressure (THP) observed by the plurality of sensors.

11. The method of claim 1, further comprising causing a computing device to generate and display a presentation including a visualization of any predicted instances of liquid loading based on the output of the liquid loading classification model.

12. A system, comprising: at least one processor;memory in electronic communication with the at least one processor; andinstructions stored in the memory, the instructions being executable by the at least one processor to: collect historical data associated with a well site, the historical data including historical data samples captured by a plurality of sensors and associated labels indicating instances of liquid loading corresponding to the historical data samples;train a liquid loading classification model based on the data samples and associated labels from the historical data;obtain input sample data, wherein obtaining the input sample data includes configuring sensor data captured by the plurality of sensors to be provided as input to the liquid loading classification model;apply the liquid loading classification model to the input sample data to generate an output including liquid loading labels indicating predicted instances of liquid loading in connection with the sensor data captured by the plurality of sensors and associated timestamps of the sensor data.

13. The system of claim 12, wherein the liquid loading classification model is a recursive neural network (RNN) model trained to predict instances of liquid loading based on series of given data samples captured by a given plurality of sensors associated with a given well site.

14. The system of claim 12, wherein the plurality of sensors includes: one or more tubing head pressure (THP) sensors;one or more casing head pressure (CHP) sensors;one or more gas rate sensors; andone or more temperature sensors.

15. The system of claim 12, wherein configuring the sensor data captured by the plurality of sensors includes pre-processing the sensor data by performing one or more of: removing outliers from the sensor data based on distances between values of samples of the sensor data and a mean value of other samples from the sensor data;removing noise from the sensor data by applying a smoothing algorithm to reduce data fluctuations within the sensor data.

16. The system of claim 15, wherein configuring the sensor data captured by the plurality of sensors further includes applying a fast-labeling model to the pre-processed sensor data to generate preliminary labels for the sensor data prior to applying the liquid loading classification model to the sampled input data.

17. The system of claim 12, wherein the liquid loading classification model is trained to predict instances of liquid loading based at least in part on a liquid loading severity metric, the liquid loading severity metric being determined based on an average of a gas rate metric and a pressure metric, and wherein the gas rate metric is based on a maximum observed gas rate over a predetermined period of time, and wherein the pressure metric is based on a difference between a case heading pressure (CHP) and a tubing head pressure (THP) observed by the plurality of sensors.

18. The system of claim 12, further comprising instructions being executable by the at least one processor to cause a computing device to generate and display a presentation including a visualization of any predicted instances of liquid loading based on the output of the liquid loading classification model.

19. A non-transitory computer readable medium storing instructions thereon, the instructions being executable by at least one processor to: collect historical data associated with a well site, the historical data including historical data samples captured by a plurality of sensors and associated labels indicating instances of liquid loading corresponding to the historical data samples;train a liquid loading classification model based on the data samples and associated labels from the historical data;obtain input sample data, wherein obtaining the input sample data includes configuring sensor data captured by the plurality of sensors to be provided as input to the liquid loading classification model;apply the liquid loading classification model to the input sample data to generate an output including liquid loading labels indicating predicted instances of liquid loading in connection with the sensor data captured by the plurality of sensors and associated timestamps of the sensor data; andcause a computing device to generate and display a presentation including a visualization of any predicted instances of liquid loading based on the output of the liquid loading classification model.

20. The non-transitory computer readable medium of claim 19, wherein the liquid loading classification model is a recursive neural network (RNN) model trained to predict instances of liquid loading based on series of given data samples captured by a given plurality of sensors associated with a given well site.