Machine-learning backbone model for fracture design

Abstract

Systems and methods presented herein are configured to generate a machine-learning (ML) backbone model for generating a fracture design output to adjust operations associated with well stimulation, fracturing, and the like. A method may include generating a synthetic physics-based dataset for hydraulic fracturing or stimulation to use machine learning or predictive modeling. The method may also include generating a machine-learning (ML) model using the synthetic physics-based dataset. The method may also include combining a real fracturing dataset with the synthetic physics-based dataset and retrain ML model using a transfer learning approach. Further, the method may include adjusting a hydraulic fracturing operation, a stimulation operation, or both, using the ML model.

Description

MACHINE-LEARNING BACKBONE MODEL FOR FRACTURE DESIGNCROSS-REFERENCE TO RELATED APPLICATION

[0001] This application claims priority to and the benefit of U.S. Provisional Patent Application Serial No. 63 / 601,367, entitled “Machine-Learning Backbone Model for Fracture Design,” filed November 21, 2023, which is hereby incorporated by reference in its entirety for all purposes.BACKGROUND

[0002] The present disclosure generally relates to systems and methods for generating a machine-learning (ML) backbone model for generating a fracture design output to adjust operations associated with well stimulation, fracturing, and the like.

[0003] This section is intended to introduce the reader to various aspects of art that may be related to various aspects of the present techniques, which are described and / or claimed below. This discussion is believed to be helpful in providing the reader with background information to facilitate a better understanding of the various aspects of the present disclosure. Accordingly, it should be understood that these statements are to be read in this light, and not as an admission of any kind.

[0004] A well stimulation job (e.g., hydraulic fracturing) consists of a design, execution, and evaluation (DEE) cycle where each stage of the job is designed for the specific well composition, each stage is executed per the design specifications, and each stage is evaluated after it is performed to ensure the quality of the job. In addition, it is common for multiple wells to bestimulated within the vicinity of other wells (known as a pad of wells), which can lead to physical influence effects of one well on another that can be detrimental to the stimulation jobs.SUMMARY

[0005] A summary of certain embodiments described herein is set forth below. It should be understood that these aspects are presented merely to provide the reader with a brief summary of these certain embodiments and that these aspects are not intended to limit the scope of this disclosure.

[0006] Certain embodiments of the present disclosure include a control system having one or more processors configured to execute processor-executable instructions stored on memory of the control system, wherein the processor-executable instructions, when executed by the one or more processors, cause the control system to initiate and implement one or more software modules in a modular manner to optimize parameters of a hydraulic stimulation job, and to provide advice regarding one or more adjustments to the parameters of the hydraulic stimulation job in substantially real-time during performance of the hydraulic stimulation job.

[0007] Various refinements of the features noted above may be undertaken in relation to various aspects of the present disclosure. Further features may also be incorporated in these various aspects as well. These refinements and additional features may exist individually or in any combination. For instance, various features discussed below in relation to one or more of the illustrated embodiments may be incorporated into any of the above-described aspects of the present disclosure alone or in any combination. The brief summary presented above is intendedto familiarize the reader with certain aspects and contexts of embodiments of the present disclosure without limitation to the claimed subject matter.BRIEF DESCRIPTION OF THE DRAWINGS

[0008] Various aspects of this disclosure may be better understood upon reading the following detailed description and upon reference to the drawings, in which:

[0009] FIG. 1 is a schematic view of at least a portion of an example implementation of a wellsite system, in accordance with embodiments of the present disclosure;

[0010] FIG. 2 is a schematic view of a portion of an example implementation of the wellsite system shown in FIG. 1, in accordance with embodiments of the present disclosure;

[0011] FIG. 3 is a schematic view of a portion of an example implementation of the wellsite system shown in FIG. 1, in accordance with embodiments of the present disclosure;

[0012] FIG. 4 shows a block diagram of a control system that may implement the disclosed techniques, in accordance with embodiments of the present disclosure;

[0013] FIG. 5 illustrates a flow diagram of a process generating and utilizing a machinelearning (ML) backbone model, in accordance with embodiments of the present disclosure;

[0014] FIG. 6 illustrates a flow diagram of one example of the process for generating a ML backbone model of FIG. 5 in accordance with embodiments of the present disclosure;

[0015] FIG. 7 illustrates a flow diagram of a process for determining one or more parameters using the process of FIG. 5 in accordance with embodiments of the present disclosure;

[0016] FIG. 8 illustrates a graph generated using the techniques of FIGS. 5 and 6, in accordance with embodiments of the present disclosure;

[0017] FIG. 9 illustrates a flow diagram of a process for determining certain outputs of FIG. 6, in accordance with embodiments of the present disclosure;

[0018] FIG. 10 illustrates a graph of perforation diameter versus gravel content, in accordance with embodiments of the present disclosure;

[0019] FIG. 11 illustrates a flow diagram of a process for determining parameters described in FIG. 6, in accordance with embodiments of the present disclosure;

[0020] FIG. 12 illustrates a graph of fraction of final concentration versus fraction of job time, in accordance with embodiments of the present disclosure;

[0021] FIG. 13 shows graphs indicating performance of the disclosed techniques for generating a model, in accordance with embodiments of the present disclosure;

[0022] FIG. 14 shows a graph illustrating a performance curve for generating a model with and without transfer learning, in accordance with embodiments of the present disclosure;

[0023] FIG. 15 shows a bar chart of the root-mean-square-error (RMSE) for various models, in accordance with embodiments of the present disclosure;

[0024] FIG. 16 shows scatterplots of partial and complete datasets, in accordance with embodiments of the present disclosure;

[0025] FIG. 17 shows a process for refining an ML backbone model, in accordance with embodiments of the present disclosure; and

[0026] FIG. 18 shows graphs indicating accuracy of predicted parameters, in accordance with embodiments of the present disclosure.DETAILED DESCRIPTION

[0027] One or more specific embodiments of the present disclosure will be described herein. These described embodiments are only examples of the presently disclosed techniques.Additionally, in an effort to provide a concise description of these embodiments, all features of an actual implementation may not be described in the specification. It should be appreciated that in the development of any such actual implementation, as in any engineering or design project, numerous implementation-specific decisions must be made to achieve the developers’ specific goals, such as compliance with system-related and business-related constraints, which may vary from one implementation to another. Moreover, it should be appreciated that such a development effort might be complex and time consuming, but would nevertheless be a routine undertaking of design, fabrication, and manufacture for those of ordinary skill having the benefit of this disclosure.

[0028] When introducing elements of various embodiments of the present disclosure, the articles “a,” “an,” and “the” are intended to mean that there are one or more of the elements. The terms “comprising,” “including,” and “having” are intended to be inclusive and mean that there may be additional elements other than the listed elements. Additionally, it should be understood that references to “one embodiment” or “an embodiment” of the present disclosure are not intended to be interpreted as excluding the existence of additional embodiments that also incorporate the recited features.

[0029] As used herein, the terms “connect,” “connection,” “connected,” “in connection with,” and “connecting” are used to mean “in direct connection with” or “in connection with via one or more elements”; and the term “set” is used to mean “one element” or “more than one element.” Further, the terms “couple,” “coupling,” “coupled,” “coupled together,” and “coupled with” are used to mean “directly coupled together” or “coupled together via one or more elements.” As used herein, the terms “up” and “down,” “uphole” and “downhole”, “upper” and “lower,” “top” and “bottom,” and other like terms indicating relative positions to a given point or element are utilized to more clearly describe some elements. Commonly, these terms relate to a reference point as the surface from which drilling operations are initiated as being the top (e.g., uphole or upper) point and the total depth along the drilling axis being the lowest (e.g., downhole or lower) point, whether the well (e.g., wellbore, borehole) is vertical, horizontal or slanted relative to the surface.

[0030] As used herein, a fracture shall be understood as one or more cracks or surfaces of breakage within rock. Fractures can enhance permeability of rocks greatly by connecting pores together and, for that reason, fractures can be induced mechanically in some reservoirs in order to boost hydrocarbon flow. Certain fractures may also be referred to as natural fractures to distinguish them from fractures induced as part of a reservoir stimulation. Fractures can also be grouped into fracture clusters (or “perf clusters”) where the fractures of a given fracture cluster (perf cluster) connect to the wellbore through a single perforated zone. As used herein, the term “fracturing” refers to the process and methods of breaking down a geological formation and creating a fracture (i.e., the rock formation around a wellbore) by pumping fluid at relatively high pressures (e.g., pressure above the determined closure pressure of the formation) in order to increase production rates from a hydrocarbon reservoir.

[0031] In addition, as used herein, the terms “real time”, “real-time”, or “substantially real time” may be used interchangeably and are intended to described operations (e.g., computing operations) that are performed without any human-perceivable interruption between operations. For example, as used herein, data relating to the systems described herein may be collected, transmitted, and / or used in control computations in “substantially real time” such that data readings, data transfers, and / or data processing steps occur once every second, once every 0.1 second, once every 0.01 second, or even more frequent, during operations of the systems (e g., while the systems are operating). In addition, as used herein, the terms “automatic” and “automated” are intended to describe operations that are performed are caused to be performed, for example, by a control system (i.e., solely by the control system, without human intervention).

[0032] Exploring, drilling, and completing hydrocarbon and other wells are relatively complicated, time consuming and, ultimately, relatively expensive endeavors. As a result, over the years, well depth and architecture have been extended in order to help enhance access to underground hydrocarbon reserves. For example, it is not uncommon to find hydrocarbon wells exceeding 30,000 feet in depth. While such well depths may increase the likelihood of accessing underground hydrocarbon reservoirs, other challenges are presented in terms of well management and the maximization of hydrocarbon recovery from such wells. For example, during completions, the well architecture may be enhanced by a series of wireline-run perforating applications tailored to introduce fractures and perforations into a formation defining the well. Thus, subsequent stimulated recovery from the reservoir may help to maximize overall production.

[0033] Machine Learning approaches have been developed for fracturing and stimulation design prediction to optimize design and production performance from the well. In general,these approaches are based on data model utilization. Physics informed machine learning (PIML) is a relatively newer branch of machine learning where a synthetic dataset is constructed based on existing knowledge of the any process which captures the underlying physics in a comprehensive way. As referred to herein, a “synthetic dataset” or “synthetic physics-based dataset” includes data that is not obtained by direct measurements. For example, the synthetic dataset may be a physical parameter within a range of physical parameters that is expected to be in a location based on the formation geology and other prior-known or predicted information. In some embodiments, the synthetic dataset may be “based on” real data, in that the synthetic dataset is extrapolated or otherwise modeled using some “real data” or a “real dataset”. As referred to here, “real data” or a “real dataset” is data that is obtained by direct measurements or converted from direct measurements. A data model is trained on this synthetic dataset which acts as a backbone of the PIML approach. Then a small dataset from a real field (actual data) is taken to optimize the predictors from the backbone model applying transfer learning. The advantage of this approach is big datasets can be created with synthetic-physics based equations / relations making small dataset from the real field viable.

[0034] The embodiments described herein include a novel, holistic approach to improving stimulation job quality though the use of PIML. There is no PIML approach present for fracturing and stimulation domain. FIG. 4 below highlights the broad workflow. In general, the techniques include determining a fluid efficiency using reservoir properties, MiniFalloff parameters, and calibration injection parameters. Further, the techniques include determining fracturing treatment properties based on the fluid efficiency, determining a fracture width, dimensionless fracture conductivity (FCD), proppant number (NproP) based on the fracturing treatment properties and the fluid efficiency. In turn, the techniques include determining a massof proppant (MproP) and / or a Pad Ratio using one or more of the fracture width, the dimensionless fracture conductivity (FCD), proppant number (Nprop). Further, the techniques include determining an amount of proppant added (PPA) (e.g., Max PPA) using proppant and perforation properties. The MprOp, the Pad Ratio, and the Max PPA may be used to determine a productivity index that may be used to predict production performance. There is also an optimizer algorithm downstream of the ML model which runs inverse solution to get the best combination of fracturing design parameters to achieve the predicted production performance.

[0035] FIG. l is a schematic view of at least a portion of an example implementation of a wellsite system 100, in accordance with embodiments of the present disclosure. FIG. 1 illustrates multiple wellbores 102 each extending from a terrain surface of a wellsite 104, a partial sectional view of a subterranean formation 106 penetrated by the wellbores 102, and various pieces of wellsite equipment or components of the wellsite system 100 located at the wellsite 104. The wellsite system 100 may facilitate recovery of oil, gas, and / or other materials that are trapped in the subterranean formation 106. In certain embodiments, each wellbore 102 may include a casing 108 secured by cement (not shown). The wellsite system 100 may be configured to transfer various materials and additives from corresponding sources to a destination location for blending or mixing and subsequent injection into one or more of the wellbores 102 during fracturing and other stimulation operations. In certain embodiments, such operations may be partially or fully automated using at least one controller, as described in greater detail herein.

[0036] In certain embodiments, the wellsite system 100 may include a mixing unit 109 (referred to hereinafter as a “mixer”) fluidly connected to one or more tanks 110 and a container 112. In certain embodiments, the container 112 may contain a first material and the tanks 110may contain a liquid. In certain embodiments, the first material may be or include a hydratable material or gelling agent, such as cellulose, clay, galactomannan, guar, polymers, synthetic polymers, and / or polysaccharides, among other examples. In addition, in certain embodiments, the liquid may be or include an aqueous fluid, such as water or an aqueous solution including water, among other examples. In certain embodiments, the mixer 109 may be configured to receive the first material and the liquid, via two or more conduits or other material transfer means (hereafter simply “conduits”) 114, 116, and mix or otherwise combine the first material and the liquid to form a base fluid, which may be or include what is known in the art as a gel. In certain embodiments, the mixer 109 may then discharge the base fluid via one or more conduits 118.

[0037] In certain embodiments, the wellsite system 100 may further include another mixer 124 fluidly connected to the mixer 109 and another container 126. In certain embodiments, the container 126 may contain a second material that may be appreciably different than the first material. For example, the second material may be or include a proppant material, such as quartz, sand, sanddike particles, silica, and / or propping agents, fibers, among other examples. In certain embodiments, the mixer 124 may be configured to receive the base fluid from the mixer 109 via the one or more conduits 118, and the second material from the container 126 via one or more conduits 128, and mix or otherwise combine the base fluid and the second material to form a mixed fluid, which may be or include what is known in the art as a fracturing fluid. In certain embodiments, the mixer 124 may then discharge the mixed fluid via one or more conduits 130.

[0038] In certain embodiments, the mixed fluid may be communicated from the mixer 124 to a common manifold 136 via the one or more conduits 130. In certain embodiments, thecommon manifold 136 may include a low-pressure distribution manifold 138, a high-pressure collection and discharge manifold 140, as well as various valves and diverters, which may be collectively configured to direct the flow of the mixed fluid in a predetermined manner. In certain embodiments, the common manifold 136 may receive the mixed fluid from the one or more conduits 130 and distribute the mixed fluid to a fleet of pump units 150 via the low- pressure distribution manifold 138. The common manifold 136 may be known in the art as a missile or a missile trailer. Although the fleet is illustrated as including four pump units 150, in other embodiments, the fleet may include other quantities of pump units 150 within the scope of the present disclosure.

[0039] Each pump unit 150 may include a pump 152, a prime mover 154, and perhaps a heat exchanger 156. In certain embodiments, each pump unit 150 may receive the mixed fluid from a corresponding outlet of the low-pressure distribution manifold 138 of the common manifold 136, via one or more conduits 142, and discharge the mixed fluid under pressure into a corresponding inlet of the high-pressure collection and discharge manifold 140 via one or more conduits 144. In certain embodiments, the mixed fluid may then be discharged from the high- pressure collection and discharge manifold 140 via one or more conduits 146.

[0040] The tanks 110, the containers 112, 126, the mixers 109, 124, the pump units 150, the manifold 136, and the conduits 114, 116, 118, 128, 130, 142, 144, 146 may collectively form a treatment (e g., stimulation) fluid system. As described herein, the treatment fluid system of the wellsite system 100 may be configured to transfer additives and produce a fracturing fluid that may be pressurized and injected into a selected wellbore 102 during hydraulic fracturing operations. However, it is to be understood that the treatment fluid system may also or instead be configured to transfer other additives and mix other treatment fluids that may be pressurizedand injected into the selected wellbore 102 during other well and / or reservoir treatment operations, such as acidizing operations, chemical injection operations, and other stimulation operations, among other examples. Accordingly, unless described otherwise, the one or more mixed fluids being produced and pressurized by the treatment fluid system for injection into a selected wellbore 102 may be referred to hereinafter simply as “a treatment fluid.”

[0041] In certain embodiments, the treatment fluid may be received by a fracturing manifold 170, which may selectively distribute the treatment fluid between the wellbores 102 via a plurality of corresponding fluid conduits 172 extending between the fracturing manifold 170 and each wellbore 102. In certain embodiments, the fracturing manifold 170 may include a plurality of remotely operated fluid flow control valves 173 (e.g., frac valves, shut-off valves), each remotely operable to fluidly connect (and disconnect) the fluid conduit 146 to (and from) a selected one or more of the fluid conduits 172 and, thus, facilitate injection of the treatment fluid into a selected one or more of the wellbores 102. The fracturing manifold 170 may be known in the art as a zipper manifold.

[0042] Each wellbore 102 may be capped by a plurality (e.g., a stack) of fluid flow control devices 174, 176, which may include or form a Christmas tree (e.g., a frac tree) including fluid flow control valves (e.g., master valves, wing valves, swab valves, etc.), spools, flow crosses (e.g., goat heads, frac heads, etc.), and fittings individually and / or collectively configured to direct and control (e.g., permit and prevent) flow of the treatment fluid into the wellbore 102 and to direct and control flow of formation fluids out of the wellbore 102. In certain embodiments, the fluid flow control valves of the fluid flow control device 174, 176 may be configured to close selected tubulars or pipes, such as the casing 108 or production tubing extending within the wellbore 102, to selectively facilitate fluid access to the wellbore 102. In certain embodiments,the fluid flow control devices 174, 176 may also include or form a blow-out preventer (BOP) stack selectively operable to prevent flow of the formation fluids out of the wellbore 102. In certain embodiments, the fluid flow control devices 174, 176 may be directly or indirectly mounted on top of a wellhead 178 (e g., tubing head adapter) terminating the wellbore 102 at the surface of the wellsite 104. In certain embodiments, each fluid flow control valve 173 of the fracturing manifold 170 may be fluidly connected to a corresponding fluid flow control device 174 via one or more fluid conduits 172, to facilitate selective fluid connection between the common manifold 136 and one or more of the wellbores 102. Thus, the fluid flow control valves 173 of the fracturing manifold 170 and the fluid flow control valves of the fluid flow control devices 174, 176 may collectively form a fluid flow control valve system configured to fluidly connect (and disconnect) one of the treatment fluid system and a pump-down system, as described herein, to (and from) a selected one or more of the wellbores 102.

[0043] In certain embodiments, a downhole intervention and / or sensor assembly, referred to herein as a tool string 180, may be conveyed within a selected one of the wellbores 102 via a conveyance line 182 operably coupled with one or more pieces of equipment at the wellsite 104. In certain embodiments, the tool string 180 may include a perforating tool configured to perforate the casing 108 and a portion of the formation 106 surrounding the wellbore 102 during perforating operations. In certain embodiments, the conveyance line 182 may be or include a cable, a wireline, a slickline, a multiline, an e-line, coiled tubing, and / or other conveyance means.

[0044] In certain embodiments, the conveyance line 182 may be operably connected to a conveyance device 184 (e.g., a wireline or coiled tubing conveyance unit) configured to apply an adjustable tension to the tool string 180 via the conveyance line 182 to convey the tool string 180through the wellbore 102. In certain embodiments, the conveyance device 184 may be or include a winch conveyance system including a reel or drum 186 storing thereon a wound length of the conveyance line 182. The drum 186 may be rotated by a rotary actuator (e.g., an electric motor, a hydraulic motor, etc.) (not shown) to selectively unwind and wind the conveyance line 182 to apply an adjustable tensile force to the tool string 180 to selectively convey the tool string 180 into and out of the wellbore 102. In certain embodiments, the conveyance line 182 may be directed, guided, and / or injected (e g., pushed downhole) into the wellbore 102 by an injection device 188 (e.g., a sheave, a pulley, a coiled tubing injector), one or more of which may be supported above the wellbore 102 via a mast, a derrick, a crane, and / or another support structure (not shown). In certain embodiments, the conveyance line 182 may include and / or be operable in conjunction with means for communication between the tool string 180, the conveyance device 184, and / or one or more other portions of the surface equipment, including a tool string control system.

[0045] The tool string 180 may be deployed into or retrieved from the wellbore 102 via the conveyance device 184 through the fluid flow control devices 174, 176, the wellhead 178, and / or a sealing and alignment assembly 189 mounted on the fluid flow control devices 174, 176 and configured to seal the conveyance line 182 during deployment, conveyance, intervention, and other wellsite operations performed via the tool string 180. The injection device 188 may, thus, guide the conveyance line 182 between the conveyance device 184 and the sealing and alignment assembly 189. In certain embodiments, the sealing and alignment assembly 189 may include a lock chamber (e.g., a lubricator, an airlock, a riser, etc.) mounted on the fluid flow control devices 174, 176, and a stuffing box configured to seal around the conveyance line 182 at the top of the lock chamber. In certain embodiments, the stuffing box may be configured to sealaround an outer surface of the conveyance line 182, such as via annular packings applied around the surface of the conveyance line 182 and / or by injecting a fluid between the outer surfaces of the conveyance line 182 and an inner wall of the stuffing box.

[0046] In certain embodiments, the sealing and alignment assembly 189 and the injection device 188 may be disconnected from above a wellbore 102 that was perforated and is now ready for stimulation (e.g., fracturing operations), and may be installed or connected above a wellbore 102 that is to be perforated in preparation for stimulation. In certain embodiments, the sealing and alignment assembly 189 and the injection device 188 may be moved from wellbore 102 to wellbore 102 and supported above a wellbore 102 by a crane or other lifting equipment. The conveyance device 184, the sealing and alignment assembly 189, the injection device 188, the tool string 180, and the conveyance line 182 may collectively form at least a portion of a perforating system configured to convey the tool string 180 (including a perforating tool) within and out of a wellbore 102 and to perforate the wellbore 102.

[0047] In certain embodiments, the wellsite system 100 may further include a pump-down system configured to inject a fluid (e.g., water) into a selected one of the wellbores 102 to perform pump-down operations to convey the tool string 180 to an intended depth along the wellbore 102. The pump-down operations may be utilized to move the tool string 180 along the wellbore 102 to facilitate wellbore plugging and perforating (“plug and perf’) operations. For example, the tool string 180 may be conveyed through the wellbore 102 to fluidly isolate an upper formation zone that has not yet been perforated from a lower formation zone that has already been perforated, and then perforate the upper formation zone. In certain embodiments, the pumping system may include a pump unit 190 configured to inject the fluid from a fluid container 194 into the selected one of the wellbores 102 containing the tool string 180 via acorresponding fluid flow control device 176 (or wellhead 178). Each pump unit 190 may include a fluid pump 192, a prime mover 193 for actuating the fluid pump 192, and perhaps a heat exchanger 195. In certain embodiments, the fluid pump 192 of the pump unit 190 may be fluidly connected to the fluid container 194 and to each fluid flow control device 176 (which may be or form a portion of the wellhead 178) via a plurality of conduits 196, which may be or form a fluid distribution manifold. In certain embodiments, pump-down and plug and perf operations may be performed in a selected wellbore 102 while stimulation operations are simultaneously performed in one or more other wellbores 102. Accordingly, when a wellbore 102 is selected to be plugged and perforated, the sealing and alignment assembly 189, the injection device 188, and the conveyance device 184 may be installed at and / or moved to the selected wellbore 102. Then, the tool string 180 may be conveyed through the wellbore 102 via the pump-down operations and utilized to perform the plug and perf operations.

[0048] In certain embodiments, the fracturing manifold 170 may include an arrangement of flow fittings and manual and remotely actuated fluid flow control valves 173, and may be configured to selectively isolate wellbores 102 by directing the treatment fluid from the common manifold 136 to a selected one or more of the wellbores 102 in which plug and perf operations have been completed and are ready to be fractured. Such operation of the fracturing manifold 170 (which may be automated or semi-automated using at least one controller, in certain embodiments) may improve the speed of transitioning between wellbores 102, and may reduce or eliminate manual adjustments, which may also reduce safety risks. Thus, the fracturing manifold 170 may be configured to facilitate “zipper” fracturing operations, which may provide improved (perhaps nearly continuous) utilization of the frac crew and equipment, resulting insubstantial improvement to the effective use of the fracturing resources and, thus, to the overall economics of the well.

[0049] In certain embodiments, the wellsite system 100 may include one or more control centers 160, each having a controller 161 (e.g., a processing device, a computer, a programmable logic controller (PLC), etc.), which may be configured to monitor and provide control to one or more portions of the wellsite system 100. The controller(s) 161 may monitor and control corresponding equipment of the treatment fluid system, the pump-down system (e.g., the pump unit 190), the plug and perf system (e.g., the conveyance device 184, the tool string 180), and the flow control valve system (e.g., the fracturing manifold 170, the fluid flow control devices 174, 176). In certain embodiments, the controller(s) 161 may be communicatively connected to the various wellsite equipment described herein, and perhaps other equipment, and may be configured to receive sensor signals from and transmit control signals to such equipment to facilitate automated or semi-automated operations described herein. For example, the controller(s) 161 may be communicatively connected to and configured to monitor and control one or more portions of the mixers 109, 124, the pump units 150, 190, the common manifold 136, the fracturing manifold 170, the fluid flow control devices 174, 176, the injection device 188, the conveyance device 184, and / or various other wellsite equipment (not shown). The controller(s) 161 may store control commands, operational parameters and set-points, coded instructions, executable programs, and other data or information, including for implementing one or more aspects of the operations described herein. Communication between the control center(s) 160 (and the controller(s) 161) and the various wellsite equipment of the wellsite system 100 may be implemented via wired and / or wireless communication means. For clarity and ease of understanding, such communication means are not depicted, and a person havingordinary skill in the art will appreciate that such communication means are within the scope of the present disclosure.

[0050] A field engineer, equipment operator, or field operator 164 (collectively referred to hereinafter as a “wellsite operator”) may operate one or more components, portions, or systems of the wellsite equipment and / or perform maintenance or repair on the wellsite equipment. For example, the wellsite operator 164 may assemble the wellsite system 100, operate the wellsite equipment (e.g., via a controller 161) to perform the stimulation operations, check equipment operating parameters, and repair or replace malfunctioning or inoperable wellsite equipment, among other operational, maintenance, and repair tasks, collectively referred to hereinafter as wellsite operations. The wellsite operator 164 may perform wellsite operations by himself or with other wellsite operators.

[0051] In certain embodiments, the controller(s) 161 may be communicatively connected to one or more human-machine interface (HMI) devices, which may be utilized by the wellsite operator(s) 164 for entering or otherwise communicating the control commands to the controller(s) 161, and for displaying or otherwise communicating information from the controller(s) 161 to the wellsite operator(s) 164. In certain embodiments, the HMI devices may include one or more input devices 167 (e.g., a keyboard, a mouse, a joystick, a touchscreen, etc.) and one or more output devices 166 (e.g., a video monitor, a printer, audio speakers, etc.). In certain embodiments, the HMI devices may also include a mobile communication device(s) 168 (e.g., a smart phone).

[0052] In certain embodiments, one or more of the containers 112, 126, 194, the mixers 109, 124, the pump units 150, 190, the fracturing manifold 170, the conveyance device 184, and thecontrol center(s) 160 may each be disposed on corresponding trucks, trailers, and / or other mobile carriers 122, 134, 198, 120, 132, 148, 197, 171, 185, 162, respectively, such as may permit their transportation to the wellsite 104. However, in certain embodiments, one or more of the containers 112, 126, 194, the mixers 109, 124, the pump units 150, 190, the fracturing manifold 170, the conveyance device 184, and the control center(s) 160 may each be skidded or otherwise stationary, and / or may be temporarily or permanently installed at the wellsite 104. In certain embodiments, the common manifold 136 and / or other equipment described herein or otherwise forming a portion of the system 100 may similarly be mobile, skidded, or otherwise installed at the wellsite 104.

[0053] FIG. 2 is a schematic view of a portion of an example implementation of the wellsite system 100 shown in FIG. 1 and indicated in FIG. 2 by reference numeral 200. The wellsite system 200 shows some of the wellsite equipment of the wellsite system 100 shown in FIG. 1, including where indicated by the same reference numerals. The following description refers to FIGS. 1 and 2, collectively.

[0054] The wellsite system 200 includes one of the wellbores 102 extending from the surface of the wellsite 104 into the rock formation 106. In certain embodiments, the wellbore 102 may be capped by the wellhead 178 terminating the wellbore 102 at the surface of the wellsite 104. In certain embodiments, the fluid flow control devices 174, 176 may be mounted on top of the wellhead 178. In certain embodiments, the fluid flow control device 174 may be fluidly connected to the fracturing manifold 170 via a corresponding conduit 172. In certain embodiments, the fluid flow control device 176 may be fluidly connected to the pump unit 190 via a corresponding conduit 196. In certain embodiments, each fluid flow control device 174, 176 may include a plurality of manually and / or remotely (e.g., electrically, pneumatically,hydraulically) operated (i.e., actuated) fluid flow control valves, each configured to selectively open and close selected tubulars or pipes, such as the casing 108 extending within the wellbore 102, to a corresponding fluid conduit 172, 196. For example, the fluid flow control device 174 may include a remotely operated fluid flow control valve 204 (e.g., a wing valve) remotely configured to fluidly connect the conduit 172 to the wellbore 102 and, thus, fluidly connect the fracturing manifold 170 to the wellbore 102. In certain embodiments, the fluid flow control device 174 may further include a remotely operated access valve 208 (e.g., swab valve) remotely configured to open top of the fluid flow control device 174 to permit vertical access to the wellbore 102 by a tool string 180. In certain embodiments, the fluid flow control device 176 may include a remotely operated fluid flow control valve 206 (e g., wing valve) remotely configured to fluidly connect the conduit 196 to the wellbore 102 and, thus, fluidly connect the pump unit 190 to the wellbore 102.

[0055] In certain embodiments, the tool string 180 may be conveyed through the wellbore 102 via a conveyance line 211 operably coupled with a winch conveyance device 210. In certain embodiments, the conveyance line 211 may be operably connected to the conveyance device 210 that is configured to apply an adjustable tension to the tool string 180 via the conveyance line 211 to convey the tool string 180 through the wellbore 102. In certain embodiments, the conveyance device 210 may be or include a winch conveyance system including a reel or drum 216 storing thereon a wound length of the conveyance line 211. In certain embodiments, the drum 216 may be rotated by a rotary actuator 217 (e.g., an electric motor, a hydraulic motor, etc.) to selectively unwind and wind the conveyance line 211 to apply an adjustable tensile force to the tool string 180 to selectively convey the tool string 180 alongthe wellbore 102. In certain embodiments, the conveyance device 210 may be carried by a truck, trailer, or another vehicle 218.

[0056] In certain embodiments, the pump unit 190 may be configured to inject a fluid (e.g., water) into each wellbore 102 via the conduits 196 to perform pump-down operations to convey the tool string 180 to an intended depth along the wellbore 102. The pump-down operations may be utilized to move the tool string 180 along the wellbore 102 to facilitate the plug and perf operations. As described herein, the tool string 180 may be conveyed through the wellbore 102 to fluidly isolate an upper portion of the wellbore 102 extending through an upper formation zone that has not yet been perforated from a lower portion of the wellbore 102 extending through a lower formation zone that has already been perforated, and then perforate the upper formation zone.

[0057] In certain embodiments, the conveyance device 210 may include a controller 212 communicatively connected to the winch device 210 and the tool string 180, such as may permit the controller 212 to receive sensor signals from and transmit control signals to such equipment to convey the tool string 180 downhole and perform various downhole operations described herein. In certain embodiments, the controller 212 may be electrically or otherwise communicatively connected to the rotary actuator 217 of the drum 216 to selectively unwind and wind the conveyance line 211 to apply an adjustable tensile force to the tool string 180 to selectively convey the tool string 180 into and out of the wellbore 102. In certain embodiments, the controller 212 may be electrically or otherwise communicatively connected to the tool string 180 via a conductor 213 extending through at least a portion of the tool string 180, through the conveyance line 211, and externally from the conveyance line 211 at the wellsite surface 104 via a rotatable joint or coupling (e.g., a collector) carried by the drum 216. In certain embodiments,the conductor 213 may transmit and / or receive electrical power, data, and / or control signals between the controller 212 and one or more portions of the tool string 180. In certain embodiments, the controller 212 may be communicatively connected to the tool string 180 and / or various portions thereof, such as various sensors and actuators of the tool string 180, via the conductor 213 to facilitate monitoring and / or control operations of the tool string 180.

[0058] The controller 212 may be communicatively connected to one or more HMI devices, which may be utilized by a wellsite operator 214 (e.g., tool string operator, winch conveyance system 210 operator) for entering or otherwise communicating control commands to the controller 212, and for displaying or otherwise communicating information from the controller 212 to the wellsite operator 214. The HMI devices may include one or more input devices 167 and one or more output devices 166. The HMI devices may also include a mobile communication device 168 carried by the wellsite operator 214.

[0059] In certain embodiments, the tool string 180 may be deployed into or retrieved from the wellbore 102 through the fluid flow control devices 174, 176, the access valve 208, and a sealing and alignment assembly 189 mounted above the access valve 208 and configured to seal the conveyance line 211 during deployment, conveyance, intervention, and other wellsite operations performed by the tool string 180. In certain embodiments, the sealing and alignment assembly 189 may include a lock chamber 220 (e.g., a lubricator, an airlock, a riser) mounted above the access valve 208, a stuffing box 222 configured to seal around the line 211 at the top of the lock chamber 220, and an injection device 224 (i.e., a pulley) configured to guide the line 211 into the stuffing box 222. In certain embodiments, a guide pulley 226 may guide the line 211 between the injection device 224 and the conveyance device 210. In certain embodiments, the stuffing box 222 may be configured to seal around an outer surface of the line 211, such asvia annular packings applied around the surface of the line 211 and / or by injecting a fluid between the outer surface of the line 211 and an inner wall of the stuffing box 222.

[0060] In certain embodiments, the conveyance line 211 may be or include a flexible conveyance line, such as a wire, a cable, a wireline, a slickline, a multiline, an e-line, and / or other conveyance means. In certain embodiments, the conveyance line 211 may include one or more metal support wires or cables configured to support the weight of the downhole tool string 180. In certain embodiments, the conveyance line 211 may also include one or more electrical and / or optical conductors 213 configured to transmit electrical energy (i.e., electrical power) and electrical and / or optical signals (e.g., information, data) therethrough, such as may permit the transmission of electrical energy, data, and / or control signals between the tool string 180 and the controller 212.

[0061] In certain embodiments, the tool string 180 may include a cable head 230 physically and / or electrically connecting the conveyance line 211 to the tool string 180, such as may permit the tool string 180 to be suspended and conveyed through the wellbore 102 via the conveyance line 211. In certain embodiments, the cable head 230 may provide telemetry and / or power distribution to the tool string 180. The tool string 180 may include at least a portion of one or more downhole devices, modules, subs, and / or other tools 232 configured to perform intended downhole operations. In certain embodiments, the tools 232 of the tool string 180 may include a telemetry / control tool, such as may facilitate communication between the tool string 180 and the controller 212 and / or control of one or more portions of the tool string 180. In certain embodiments, the telemetry / control tool may include a downhole controller (not shown) communicatively connected to the controller 212 via the conductor 213 and to other portions of the tool string 180. In certain embodiments, the tools 232 of the tool string 180 may furtherinclude one or more inclination and / or directional sensors, such as one or more accelerometers, magnetometers, gyroscopic sensors (e.g., micro-electro-mechanical system (MEMS) gyros), and / or other sensors for determining the orientation and / or direction of the tool string 180 within the wellbore 102. In certain embodiments, the tools 232 of the tool string 180 may also include a depth correlation tool, such as a casing collar locator (CCL) for detecting ends of casing collars by sensing a magnetic irregularity caused by the relatively high mass of an end of a collar of the casing 108. In certain embodiments, the depth correlation tool may also or instead be or include a gamma ray (GR) tool that may be utilized for depth correlation.

[0062] In certain embodiments, the tool string 180 may also include one or more perforating guns or tools 234 configured to perforate or form holes though the casing 108, the cement, and the portion of the formation 106 surrounding the wellbore 102 to prepare the well for fracturing. In certain embodiments, each perforating tool 234 may contain one or more shaped explosive charges operable to perforate the casing 108, the cement, and the formation 106 upon detonation. In certain embodiments, the tool string 180 may also include a plug 236 and a plug setting tool 238 that, when activated, sets the plug 236 at a predetermined position within the wellbore 102, such as to isolate or seal an upper portion (e g., zone) of the wellbore 102 from a lower portion (e.g., zone) of the wellbore 102 and, in certain embodiments, disconnects the borehole assembly (BHA) from the plug 236. The plug 236 may be permanent or retrievable, facilitating the lower portion (e.g., zone) of the wellbore 102 to be permanently or temporarily isolated or sealed from the upper portion (e.g., zone) of the wellbore 102 before perforating operations.

[0063] In certain embodiments, the treatment fluid system may further include a control center 250 containing a controller 252 (e.g., a processing device, a computer, a PLC, etc.), which may be configured to monitor and provide control to one or more portions of the treatment fluidsystem. The controller 252 may be communicatively connected to the various equipment of the treatment fluid system and may be configured to receive sensor signals from and transmit control signals to such equipment to facilitate automated or semi-automated operations described herein. For example, the controller 252 may be communicatively connected to and configured to monitor and control one or more portions of the mixers 109, 124, the pump units 150, the common manifold 136, and / or various other wellsite equipment (not shown). The controller 252 may store control commands, operational parameters and set-points, coded instructions, executable programs, and other data or information, including for implementing one or more aspects of the operations described herein. Communication between the control center 250 (and the controller 252) and the various equipment of the treatment fluid system may be implemented via wired and / or wireless communication means. For clarity and ease of understanding, such communication means are not depicted, and a person having ordinary skill in the art will appreciate that such communication means are within the scope of the present disclosure.

[0064] In certain embodiments, the controller 252 may be communicatively connected to one or more HMI devices, which may be utilized by a wellsite operator 254 (e.g., fracturing system operator) for entering or otherwise communicating control commands to the controller 252, and for displaying or otherwise communicating information from the controller 252 to the wellsite operator 254. In certain embodiments, the HMI devices may include one or more input devices 167 and one or more output devices 166. In addition, in certain embodiments, the HMI devices may also include a mobile communication device 168 carried by the wellsite operator 254.

[0065] In certain embodiments, the pump-down system may further include a controller 262(e.g., a processing device, a computer, a PLC, etc.) disposed in association with the pump unit190 and / or fluid container 194. The controller 262 may be configured to monitor and providecontrol to one or more portions of the pump-down system. The controller 262 may be communicatively connected to the various equipment of the pump-down system and may be configured to receive sensor signals from and transmit control signals to such equipment to facilitate automated or semi-automated operations described herein. For example, the controller 262 may be communicatively connected to and configured to monitor and control one or more portions of the pump unit 190, the fluid container 194, and / or various other wellsite equipment (not shown). The controller 262 may store control commands, operational parameters and set -points, coded instructions, executable programs, and other data or information, including for implementing one or more aspects of the operations described herein. Communication between the controller 262 and the equipment of the pump-down system may be implemented via wired and / or wireless communication means. For clarity and ease of understanding, such communication means are not depicted, and a person having ordinary skill in the art will appreciate that such communication means are within the scope of the present disclosure.

[0066] In certain embodiments, the controller 262 may be communicatively connected to one or more HMI devices, which may be utilized by a wellsite operator 264 (e.g., pump-down operator) for entering or otherwise communicating control commands to the controller 262, and for displaying or otherwise communicating information from the controller 262 to the wellsite operator 264. In certain embodiments, the HMI devices may include one or more input devices 167 and one or more output devices 166. In addition, in certain embodiments, the HMI devices may also include a mobile communication device 168 carried by the wellsite operator 264.

[0067] In certain embodiments, the wellsite systems 100, 200 may further include a central controller 272 (e g., a processing device, a computer, a PLC, etc.) configured to monitor andprovide control to one or more portions of the wellsite systems 100, 200. The controller 272 may store control commands, operational parameters and set-points, coded instructions, executable programs, and other data or information, including for implementing one or more aspects of the operations described herein. The controller 272 may be communicatively connected to the various equipment of the wellsite systems 100, 200 and may be configured to receive sensor signals from and transmit control signals to such equipment to facilitate automated or semi -automated operations described herein. For example, in certain embodiments, the controller 272 may be communicatively connected to the controller 212 and configured to monitor and control one or more portions of the plug and perf system (e.g., the conveyance device 210, the tool string 180) via the controller 212. In addition, in certain embodiments, the controller 272 may be further communicatively connected to the controller 252 and configured to monitor and control one or more portions of the treatment fluid system (e.g., the mixers 109, 124, the pump units 150) via the controller 252. In addition, in certain embodiments, the controller 272 may be further communicatively connected to the controller 262 and configured to monitor and control one or more portions of the pump-down system (e.g., the pump unit 190, the fluid container 194) via the controller 262. In addition, in certain embodiments, the controller 272 may be further communicatively connected to the fluid flow control devices 174, 176 (e.g., the fluid flow control valves 204, 206) and the access valve 208 associated with each wellbore 102 and the fracturing manifold 170 (e.g., fluid flow control valves 173), such as may permit the controller 272 to monitor and control the fluid flow control devices 174, 176, the access valves 208, and the fracturing manifold 170. The controller 272 may, thus, monitor and / or control injection of treatment fluid via the fluid flow control device174 and injection of water or other fluid via the fluid flow control device 176 into one or more selected wellbores 102.

[0068] Communication between the controller 272 and the controllers 212, 252, 262, the fluid flow control devices 174, 176, the access valves 208, and the fracturing manifold 170 may be implemented via wired and / or wireless communication network 276 (e.g., a local area network (LAN), a wide area network (WAN), the internet, etc.). For clarity and ease of understanding, details of such communication means are not depicted, and a person having ordinary skill in the art will appreciate that such communication means are within the scope of the present disclosure.

[0069] In certain embodiments, the controller 272 may be communicatively connected to one or more HMI devices, which may be utilized by a wellsite operator 274 for entering or otherwise communicating control commands to the controller 272, and for displaying or otherwise communicating information from the controller 272 to the wellsite operator 274. In certain embodiments, the HMI devices may include one or more input devices 167 and one or more output devices 166. In addition, in certain embodiments, the HMI devices may also include a mobile communication device 168 carried by the wellsite operator 274. In certain embodiments, the controller 272, the HMI devices 166, 167, and the wellsite operator 274 may be located at the wellsite surface 104. For example, the controller 272 may be installed or housed in a control center (e g., a facility, a trailer, etc.) housing one of the other controllers 212, 252, 262. However, the controller 272, the HMI devices 166, 167, and the wellsite operator 274 may also or instead be located off-site (e.g., a data center) at a distance from the wellsite surface 104.

[0070] As described herein, the central controller 272 and / or the wellsite operator 274 using the central controller 272 may monitor and provide control to one or more portions of the wellsite systems 100, 200 via direct communication with selected wellsite equipment and / or indirect communication with selected wellsite equipment via dedicated equipment controllers 212, 252, 262 for controlling such wellsite equipment. For example, during pump-down operations, after the tool string 180 is made up and positioned within a selected one of the wellbores 102 below the wellhead 178, the controller 272 and / or the wellsite operator 274 using the controller 272 may initialize operation of the pump unit 190 to pump a fluid (e.g., water) from the fluid container 194. The controller 272 and / or the wellsite operator 274 may also cause the remotely operated fluid valve 206 of the fluid flow control device 176 to open to permit the fluid to be injected into the wellbore 102 containing the tool string 180. The fluid may be injected into the wellbore 102 when the tool string 180 is conveyed within a vertical portion of the wellbore 102 just below the fluid flow control device 176 or when the tool string 180 stops descending within the wellbore 102 by way of gravity. The fluid injected into the wellbore 102 may flow downhole, as indicated by arrows 240, thereby forming an increased pressure zone behind (i.e., uphole from) the tool string 180 that is greater than fluid pressure in front of (i.e., downhole from) the tool string 180. Such pressure differential may push or otherwise impart a downhole-directed force operable to move the tool string 180 in the downhole direction. The fluid flowing downhole 240 may also or instead cause friction or drag while the fluid flows around or past the tool string 180, as indicated by arrows 242. The friction may drag or otherwise impart a downhole-directed force operable to move the tool string 180 in the downhole direction. During the pump-down operations, the fluid passing 242 the tool string 180 may escape from the wellbore 102 into the formation 106 in front of the tool string 180 viapreviously made perforations 107, as indicated by arrows 244, thereby permitting the fluid pumped into the wellbore 102 to continually flow around or past the tool string 180 until the tool string 180 is conveyed to an intended depth within the wellbore 102.

[0071] In certain embodiments, while the fluid is being injected into the wellbore 102 by the fluid pump unit 190 during the pump-down operations, the controller 272 and / or the wellsite operator 274 may operate the conveyance device 210 to selectively rotate the drum 216 to unwind the conveyance line 211 to permit the pumped fluid to move the tool string 180 downward along the wellbore 102 at an intended speed and to an intended depth. In certain embodiments, after the tool string reaches the intended depth, the controller 272 and / or the wellsite operator 274 may shut off the pump unit 190 and close the fluid flow control valve 206.

[0072] In certain embodiments, while the fluid is being injected into the wellbore 102 by the fluid pump unit 190 during the pump-down operations, the controller 272 and / or the wellsite operator 274 may also operate the treatment fluid system to mix and pump the treatment fluid, open the fluid flow control valve 204 of the fluid flow control device 174, and operate a corresponding fluid flow control valve 173 of the fracturing manifold 170 of one or more of the other wellbores 102 not undergoing the pump-down operations to direct the treatment fluid therein.

[0073] In certain embodiments, after the plug and perf operations of the wellbore 102 are complete, the controller 272 and / or the wellsite operator 274 may operate the conveyance device 210 to pull the tool string 180 out of the wellbore 102 through the fluid flow control devices 174, 176 and close the access valve 208. Thereafter, the controller 272 and / or the wellsite operator 274 may operate the treatment fluid system to mix and pump the treatment fluid, open the fluidflow control valve 204 of the fluid flow control device 174, and operate a corresponding fluid flow control valve 173 of the fracturing manifold 170 to direct the treatment fluid into the newly perforated wellbore 102.

[0074] FIG. 3 is a schematic view of a portion of an example implementation of the wellsite system 100 shown in FIG. 1 and indicated in FIG. 3 by reference numeral 300. The wellsite system 300 shows some of the wellsite equipment of the wellsite systems 100, 200 shown in FIGS. 1 and 2, respectively, including where indicated by the same reference numerals. The following description refers to FIGS. 1 and 3, collectively.

[0075] The wellsite system 300 includes one of the wellbores 102 extending from the surface of the wellsite 104 into the formation 106. In certain embodiments, the wellbore 102 may be capped by the wellhead 178 terminating the wellbore 102 at the surface of the wellsite 104. In certain embodiments, the fluid flow control devices 174, 176 may be mounted on top of the wellhead 178. In certain embodiments, the fluid flow control device 174 may be fluidly connected to the fracturing manifold 170 via a corresponding conduit 172. In certain embodiments, the fluid flow control device 176 may be fluidly connected to the pump unit 190 via a corresponding conduit 196.

[0076] In certain embodiments, the tool string 180 may be conveyed through the wellbore 102 via a conveyance line 311 carried by a line storage device 310, which may include a reel or drum 316 storing thereon a wound length of the conveyance line 31 1. In certain embodiments, the drum 316 may be rotated by a rotary actuator 317 (e.g., an electric motor, a hydraulic motor, etc.) to selectively unwind and wind the conveyance line 311. In certain embodiments, the line storage device 310 may be carried by a truck, trailer, or another vehicle 318.

[0077] In certain embodiments, the tool string 180 may be deployed into or retrieved from the wellbore 102 through the fluid flow control devices 174, 176, the access valve 208, and the sealing and alignment assembly 189 that is mounted above the access valve 208 and configured to seal the conveyance line 311 during deployment, conveyance, intervention, and other wellsite operations performed by the tool string 180. In certain embodiments, the sealing and alignment assembly 189 may include a lock chamber 220 (e.g., a lubricator, an airlock, a riser) mounted above the access valve 208, a stuffing box 222 configured to seal around the line 311 at the top of the lock chamber 220, and an injection device 324 (i.e., coiled tubing injector) configured to guide the line 311 into the stuffing box 222. In certain embodiments, the stuffing box 222 may be configured to seal around an outer surface of the line 311, such as via annular packings applied around the surface of the line 311 and / or by injecting a fluid between the outer surface of the line 311 and an inner wall of the stuffing box 222.

[0078] In certain embodiments, the conveyance line 311 may be or include coiled tubing. In certain embodiments, the conveyance line 311 may include or contain one or more electrical and / or optical conductors 313 configured to transmit electrical energy (i.e., electrical power) and electrical and / or optical signals (e.g., information, data), such as may permit the transmission of electrical energy, data, and / or control signals between the tool string 180 and the line storage device 310.

[0079] In certain embodiments, the injection device 324 may be or include an injector head 326 configured to run and retrieve the line 311 into and out of the wellbore 102. In certain embodiments, a gooseneck 328 may be mounted on top of the injector head 326 to feed or direct a line 311 around a controlled radius into the injector head 326. In certain embodiments, the injector head 326 may include opposing circulating members, such as may be configured tocompress or otherwise grip the line 311 to support the weight of the downhole tool string 180 within the wellbore 102. For example, the injector head 326 may be a belt-type injector head including a pair of opposing belts 330 circulated by upper and lower rollers 332, 334. In certain embodiments, a corresponding set of cylinders 336 may push each belt 330 against the line 311 to maintain a sufficient pressure and, thus, friction between the belts 330 and an outer surface of the line 311 to grip the line 311. In certain embodiments, the belts 330 may include rubber, such as EPDM (ethylene propylene diene monomer). However, other embodiments of the injector head 326 may include chains instead of the belts 330. In certain embodiments, the injector head 326 may be mounted to or otherwise above the stuffing box 222 configured to fluidly seal against the line 311 while it exits or enters the injector head 326.

[0080] One or more of the rollers 332, 334 may be operated by a corresponding motor 338 mechanically connected to the rollers 332, 334. A gear box or transmission (not shown) may be mechanically or otherwise operatively connected between each motor 338 and the corresponding rollers 332, 334, such as may facilitate control of rotational speed and torque applied to the rollers 332, 334. When the motors 338 are implemented as hydraulic motors, a pump may be driven by an engine or an electric motor (neither shown) to supply hydraulic energy. The hydraulics system may provide variable speed commands. When the motors 338 are implemented as electrical motors, the motors 338 may be electrically connected to an electrical motor controller (e.g., a variable frequency drive, a chopper) (not shown) configured to control the speed and / or torque of the motors 338, such as by controlling the frequency and / or the amplitude of the electrical energy supplied to the motors 338. Although the injector head 326 is shown mounted above the lock chamber 220 and the stuffing box 222, the injector head 326 maybe installed or otherwise disposed within the pressure contained volume of the lock chamber220, below the stuffing box 222.

[0081] In certain embodiments, the line storage device 310 and the injection device 324 may include or be associated with a controller 312 communicatively connected to the line storage device 310 and the injection device 324, such as may permit the controller 312 to receive sensor signals from and transmit control signals to such equipment to perform various downhole operations described herein. In certain embodiments, the controller 312 may be electrically or otherwise communicatively connected to the rotary actuator 317 of the drum 316 and to the motors 338 of injection device 324 to selectively unwind and wind the conveyance line 311 to apply an adjustable compressive and tensile force to the line 311 to selectively convey the tool string 180 into and out of the wellbore 102. In certain embodiments, the controller 312 may be electrically or otherwise communicatively connected to the tool string 180 via a conductor 313 extending through at least a portion of the tool string 180, through the conveyance line 31 1, and externally from the conveyance line 311 at the wellsite surface 104 via a rotatable joint or coupling (e.g., a collector) carried by the drum 316. In certain embodiments, the conductor 313 may transmit and / or receive electrical power, data, and / or control signals between the controller 312 and one or more portions of the tool string 180. In certain embodiments, the controller 312 may be communicatively connected to the tool string 180 and / or various portions thereof, such as various sensors and actuators of the tool string 180, via the conductor 313 to facilitate monitoring and / or control operations of the tool string 180.

[0082] In certain embodiments, the controller 312 may be communicatively connected to one or more HMI devices, which may be utilized by a wellsite operator 314 (e.g., tool string operator, coiled tubing system operator, injector head operator) for entering or otherwisecommunicating control commands to the controller 312, and for displaying or otherwise communicating information from the controller 312 to the wellsite operator 314. In certain embodiments, the HMI devices may include one or more input devices 167 and one or more output devices 166. In addition, in certain embodiments, the HMI devices may also include a mobile communication device 168 carried by the wellsite operator 314.

[0083] In certain embodiments, the wellsite systems 100, 300 may further include a central controller 272 configured to monitor and provide control to one or more portions of the wellsite systems 100, 300. The controller 272 may be communicatively connected to the various equipment of the wellsite systems 100, 300 and may be configured to receive sensor signals from and transmit control signals to such equipment to facilitate automated or semi-automated operations described herein. For example, in certain embodiments, the controller 272 may be communicatively connected to the controller 312 and configured to monitor and control one or more portions of the plug and perf system (e.g., the line storage device 310, the injector head 324, and the tool string 180) via the controller 312. In addition, in certain embodiments, the controller 272 may be further communicatively connected to the controller 252 and configured to monitor and control one or more portions of the treatment fluid system via the controller 252. In addition, in certain embodiments, the controller 272 may be further communicatively connected to the controller 262 and configured to monitor and control one or more portions of the pump-down system (e.g., the pump unit 190, the fluid container 194) via the controller 262. In addition, in certain embodiments, the controller 272 may also be communicatively connected to the fluid flow control devices 174, 176 (e.g., the fluid flow control valves 204, 206) and access valve 208 associated with each wellbore 102 and with the fracturing manifold 170 (e.g., fluid flow control valves 173), such as may permit the controller 272 to monitor and control the fluidflow control devices 174, 176 and the fracturing manifold 170. The controller 272 may, thus, monitor and / or control injection of treatment fluid via the fluid flow control device 174 and injection of water or other fluid via the fluid flow control device 176 into one or more selected wellbores 102.

[0084] Communication between the controller 272 and the controllers 312, 252, 262, the fluid flow control devices 174, 176, the access valve 208, and the fracturing manifold 170 may be implemented via wired and / or wireless communication network 276 (e.g., a local area network (LAN), a wide area network (WAN), the internet, etc.). For clarity and ease of understanding, details of such communication means are not depicted, and a person having ordinary skill in the art will appreciate that such communication means are within the scope of the present disclosure.

[0085] As described herein, the central controller 272 and / or the wellsite operator 274 using the central controller 272 may be configured to monitor and provide control to one or more portions of the wellsite systems 100, 300 via direct communication with wellsite equipment and / or indirect communication with wellsite equipment via dedicated equipment controllers 312, 252, 262 for controlling corresponding wellsite equipment.

[0086] In certain embodiments, the line storage device 310 and the injector head 324 may be collectively operated by the central controller 272 and / or the wellsite operator 274 using the central controller 272 to convey the tool string 180 through the wellbore 102 without pumping the fluid into the wellbore 102. In certain embodiments, the conveyance line 311 may be sufficiently rigid to permit conveyance of the tool string 180 to an intended depth along the wellbore 102, including in a deviated or horizontal portion of the wellbore 102. During suchconveyance operations, the central controller 272 and / or the wellsite operator 274 using the central controller 272 may operate the line storage device 310 to selectively rotate the drum 316 to unwind the conveyance line 311 and to inject the conveyance line 311 into the wellbore 102 via the injection device 324 to push or otherwise move the tool string 180 downhole along the wellbore 102 at an intended speed and to an intended depth.

[0087] In certain embodiments, while the plug and perf operations of the wellbore 102 are being performed, the controller 272 and / or the wellsite operator 274 may also operate the treatment fluid system to mix and pump the treatment fluid, open the fluid flow control valve 204 of the fluid flow control device 174, and operate a corresponding fluid flow control valve 173 of the fracturing manifold 170 of one or more of the other wellbores 102 not undergoing the plug and perf operations to direct the treatment fluid therein. In certain embodiments, after the plug and perf operations of the wellbore 102 are complete, the central controller 272 and / or the wellsite operator 274 using the central controller 272 may operate the fracturing manifold 170 to direct the treatment fluid into such wellbore 102 and / or operate the fluid access valve 204 of the fluid flow control device 174 associated with such wellbore 102 to permit the treatment fluid to be injected into the newly perforated wellbore 102.

[0088] As mentioned herein, there is no PIML approach present for fracturing and stimulation domain. FIG. 5 below highlights the broad workflow. The workflow described herein may begin with a robust synthetically constructed dataset that captures all or the most relevant portion of the physics of the fracturing and stimulation process. This step is described in further detail with respect to FIGS. 13 and 14. In general, the variables may be split into predictors and outputs to train an ML model to create a backbone model (e.g., ML backbone model) from this synthetic dataset. In general, any amount of data may be generated syntheticallyin accordance with the disclosed techniques. The ML backbone model predicts the fracturing design parameters and production output with the given set of inputs (detailed later). The next component of the ML backbone model is the transfer learning where the real field data is taken and concatenated with the synthetic dataset to rerun the model and update the cost functions (of the backbone model) based on the real field data. There is also an optimizer algorithm downstream of the ML backbone model which runs inverse solution to get the best combination of fracturing design parameters to achieve the predicted production performance.

[0089] FIG. 4 illustrates a stimulation control system 350 that may perform the operations described herein. The stimulation control system 350 includes one or more processors 352, one or more storage media 354

[0090] In certain embodiments, the one or more processors 352 may include a microprocessor, a microcontroller, a processor module or subsystem, a programmable integrated circuit, a programmable gate array, a digital signal processor (DSP), or another control or computing device. In certain embodiments, the one or more storage media 354 may be implemented as one or more non-transitory computer-readable or machine-readable storage media. In certain embodiments, the one or more storage media 354 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); or other types of storage devices. Note that the computer-executable instructions described herein be provided on one computer-readable or machine-readable storage medium of the storage media354, or alternatively, may be provided on multiple computer-readable or machine-readable storage media distributed in a large system having possibly plural nodes. Such computer- readable or machine-readable storage medium or media are considered to be part of an article (or article of manufacture), which may refer to any manufactured single component or multiple components. In certain embodiments, the one or more storage media 354 may be located either in the machine running the machine-readable instructions, or may be located at a remote site from which machine-readable instructions may be downloaded over a network for execution.

[0091] In certain embodiments, the processor(s) 352 may be connected to a network interface 356 of the well stimulation control system 350 to allow the well stimulation control system 350 to communicate with the various components, actuators, flow control devices, and other equipment (e.g., surface equipment, downhole equipment, or both) described herein. For example, and as shown, the well stimulation control system 350 is coupled to the controllers 272 and the flow control devices 174, 176.

[0092] It should be appreciated that the well stimulation control system 350 illustrated in FIG. 4 is only one example of a well stimulation control system, and that the well stimulation control system 350 may have more or fewer components than shown, may combine additional components not depicted in the embodiment of FIG. 4, and / or the well stimulation control system 350 may have a different configuration or arrangement of the components depicted in FIG. 4. In addition, the various components illustrated in FIG. 4 may be implemented in hardware, software, or a combination of both hardware and software, including one or more signal processing and / or application specific integrated circuits. Furthermore, the operations of the well stimulation control system 350 as described herein may be implemented by running one or more functional modules in an information processing apparatus such as application specificchips, such as application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), programmable logic devices (PLDs), systems on a chip (SOCs), or other appropriate devices. These modules, combinations of these modules, and / or their combination with hardware are all included within the scope of the embodiments described herein.

[0093] FIG. 5 illustrates a flow diagram of a process 410 for predicting production performance, in accordance with embodiments of the present disclosure. The process 410 may be performed by one or more processors. For example, the process 410 in FIG. 6, and other processes described herein, may be performed using one or more processors 352 of the well stimulation control system 350, or any suitable processor. In general, the process 410 includes generating a physic-based dataset construction (block 412). The physics-based dataset construction may include predictors and outputs. The process 410 includes splitting the predictors and outputs (block 414). Further, the process 410 includes training a machinelearning (ML) model (block 416), which may be synthetic ML model. The ML model may receive (e.g., be introduced to) small datasets from real fields (block 418). In turn, the process 410 may include optimizing the cost functions for the predictors split at block 414 (block 420). Then, the original ML model is retrained (block 422). The retrained model may be used to predict fracturing treatment parameters for new wells (block 424). In turn, a processor may utilize the disclosed model to predict production performance (block 426). The processor may then utilize the disclosed model to adjust one or more oil and gas operations based on the predicted production performance. For example, the processor (e.g., a control system including one or more processors) may adjust one or more of the components or equipment described with respect to FIGS. 1-3.). In some embodiments, the processor may perform an injection diagnostic test using the model.

[0094] A specific non-limiting example of the process 410 is shown in FIG. 6. In particular, FIG. 6 illustrates a flow diagram of a method 440 for generating a synthetic dataset, in accordance with embodiments of the present disclosure. In general, FIG. 6 shows a specific example of an implementation of the process of FIG. 5 is shown in FIG. 6, illustrated the framework of the synthetic dataset construction. In general, the process 440 includes receiving input reservoir properties (block 442). Then, the process 440 includes generating, at block 444, input parameters for the MiniFalloff (MFO) or the injectivity test, such as pump time, net pressure generated with brine, etc. then the techniques may include using calculations for the MFO evaluation results 446 based on physical correlations. Then, the techniques described with respect to blocks 442 and 444 may be performed in a generally similar manner, at block 448, to calculate calibration injection inputs, at block 450. Using the calibration injection outputs and the MiniFalloff evaluation results, the fluid efficiency (F.E.) (e.g., crosslinked fluid efficiency) of the fracturing fluid is calculated (block 452). In some embodiments, the calibration injection diagnostic test may be done based on fracturing fluid (typically crosslinked).

[0095] Further, the process 440 may include receiving, at block 456, fracturing treatment inputs such as fracture (proppant pack) permeability, fracture half length, fracture width, to calculate or determine outputs such as dimensionless fracture conductivity (FCD), proppant number (Nprop) etc. (block 458). This loop gives is the proppant total mass and the pad ratio (also known as pad fraction or pad percent).

[0096] Further still, the process 440 includes inputting proppant properties 460 to determine a first maximal amount of proppant added (PPA) for bridging in a fracture (block 462).Further, the process 440 includes inputting perforation properties 464 such as proppant diameter, perforation diameter, etc. (at block 460) to determine a second maximal PPA for bridging at aperforation. Then, the process 440 includes determining, at block 468, the minimal proppant concentration output using the first and second maximal PPA determined at blocks 462 and 466. Then, the process 440 includes determining dimensionless productivity index (JD) using the outputs of blocks 458 and 468. JD may be used to determine a production output (e.g., a fracture design output) that controls operation of the equipment described in FIG. 2, generate a display on a GUI, or otherwise inform oil and gas operations.

[0097] As described herein, the techniques include using a synthetic dataset based on basic reservoir and rock parameters to make the calculations fully integrated and dependent. In general, FIG. 7 shows example calculations of fracture compliance, bottomhole static temperature (BHST), brine filtrate viscosity, formation compressibility, and total compressibility. In particular, FIG. 7 shows an example of randomizing some (e.g., one or more) of the parameters and calculate the dependent properties based on classical techniques and correlations. It should be noted that the initial randomization ranges for each variable could easily be extended under this approach to the required values for any basin. For the initial dataset, various practical ranges may be used in the dataset, several of which are shown as the “randomized range”. To calculate the formation compressibility using for Newman’s correlation, correlations corresponding to limestone and sandstone may be used. For Shale and other unconventional formations (e.g., volcanic rocks such as Basalt, Granite) the typical values of formation compressibility can simply be used directly, randomized within the appropriate ranges.

[0098] The description below relates to the use of the physics of the fracturing process to generate to the outputs described with respect to blocks 458 and 468 in FIG. 6. FIG. 8 andFIG. 9 show one of the outputs set, which is the F.E. of fracturing fluid (e.g., as described inblock 452 of FIG. 6) and different Pad Ratio estimates (e.g., as described in block 458 of FIG. 6) based on an end user design objective.

[0099] FIG. 8 illustrates a graph 500 of spurt loss (y-axis) versus kAP (md psi) for various fluid systems (e.g., linear HPG, Borate XL HPG, and Titanate XL HPG) in accordance with embodiments of the present disclosure. In some embodiments, the pumping time, net pressure (pnet), G-time at closure (Gc) during the MiniFalloff pumping are randomized first to calculate fluid leakoff of brine / water. Fracture compliance (i.e. fracture height / shear modulus) and rpare known rock properties which are randomized to then calculate CL, which can be calculated from fluid (brine) used in MFO using below correlation:

[0100] To make sure that CL and consequently F.E. obtained from above process are consistent to reservoir properties, CL values are tailored to leakoff controlled by fluid penetration / filtration into the formation, Cf, and leakoff controlled by the resistance to displace reservoir fluids, CR, based on the correlation shown below:where the pumping time (MFO), Tp, may be between 5 to 20 minutes, the fracture length, Lp, may be between 10 to 20 ft, and the net pressure may be between 50-500 psi, APR / APT is randomized between three conditions, i.e., 1.2 and 1.5, 1, Cf / Cr. Based on this a dominant leakoff mechanism is determined. Put differently, the three conditions may be when APR / APT > 1, APR / APT = 1, or APR / APT < 0.7. Note that dimensionless time Tpare introduced to calculate CR with a randomized fracture length for MFO injection.

[0101] Then a processor may determine or check if crosslinked (XL) gel filter cake can provide any smaller leakoff using below:fWhere bottomhole static temperature (BHST) is in degrees Fahrenheit and for 2% KC1.CL-XL —min( < Cw (8)Next step is to estimate the F.E. of main fracturing treatment from the MFO parameters. Use CL-XL and Pnet (net pressure from MFO), APnet (pressure change compared to MFO), tpcan be randomized and below correlations are used along with FIG. 8.

[0102] Gcfor XL is determined from:

[0103] FIG. 9 illustrates a flow diagram of a method 510 for determining certain properties generally described with respect to FIGS. 13 and 14, in accordance with embodiments of the present disclosure. For example, FIG. 9 shows the output of F.E. and multiple pad ratios that may be outputs of the disclosed model.

[0104] FIG. 10 illustrates a graph 520 of perforation diameter versus gravel content, in accordance with embodiments of the present disclosure. After comprehensively calculating all the parameters for F.E. next step is to get the proppant related outputs. FIG. 11 shows the detailed calculation workflow along with the correlations. Starting with the Pnet and fracture properties, which are randomized, fracture width and FCD are calculated. After this, the correlations may be used to calculate the Nprop, VproPand from that the total mass of proppant and JD as per the correlations detailed below:

[0105] Then finally, based on the mean diameter of the proppant and the perforation, bridging parameters are calculated to evaluate what will be the maximum proppant concentration based on the bridging occurring in the fracture or the perforation. The bridging condition for the fracture width is given below, where bridging factor, b = 2.5, dis the particle diameter, and Cpis the particle volume fraction in flowing suspension:

[0106] And the perforation bridging criteria is described by the empirical curve shown below and Ib / gal is capped at 30.0 for the dataset. Accordingly, the maximum proppant concentration may be utilized to generate a fracturing treatment output. The fracturing treatment output may include a control signal that causes certain injection operations to occur.

[0107] FIG. 11 illustrates a flow diagram of a process 530 for determining one or more of the parameters described with respect to FIGS. 12-14 and 16, in accordance with embodiments of the present disclosure. For example, the process 530 illustrates randomized inputs that may be used to determine a fracture width, FCD (e.g., block 456 of FIG. 6), NproP(e.g., block 456 of FIG. 6), Vprop, Prop density bulk (e g., block 460 of FIG. 6), slurry fraction, perforation diameter (e.g., block 464 of FIG. 6), PPA of a perforation for bridging (PPA_perf bridge) (e.g., block 466 of FIG. 6), and PPA for bridging in fracture (PPA frac bridge) (e.g., block 462 of FIG. 6).

[0108] Also, a pumping schedule can be generated by F.E. and PPA Final to have fairly uniform proppant concentration at the end of treatment, referring to the chart in FIG. 12. FIG 12 illustrates a graph 600 of fraction of final concentration (y-axis) versus fraction of job time (x-axis), in accordance with embodiments of the present disclosure.

[0109] After the synthetic dataset is completed, it is split in predictors and outputs. The Appendices detail example variables with the split between randomized, calculated and inputs, outputs etc. A comparative ML workflow can be established as the conclusive step to create the ML backbone model, which can now be used directly for design predictions (based on physical and numerical correlations) or ready for the transfer learning module with the real dataset. FIG. 13 shows the bar graphs indicating results from the backbone model withpromising results with multiple regressors as well as with the Neural Network (NN) approach. FIG. 13 also shows scatter plots indicating performance of the disclosed techniques for generating a model (e.g., ML backbone model), in accordance with embodiments of the present disclosure.

[0110] Once the transfer learning is performed, the performance curve completely changes compared to a typical data model deep learning / machine learning workflow. The highlights are depicted in FIG. 14. The starting point and rate of learning is higher because of the existing knowledge from previous dataset and that also leads to the final converged skill to be higher compared to the data model. FIG. 14 shows a graph 620 illustrating a performance curve for generating a model with and without transfer learning, in accordance with embodiments of the present disclosure. Performance curve of transfer learning showing the benefits of learning from the existing dataset (physics based synthetic dataset in our case) due to prior learning experience.

[0111] When used on a small basin dataset from open literature with 70 fracturing treatments data, it is observed that the impact of the transfer learning approach as depicted in FIG. 15 shows lower error rates relative to the pure data model approach. FIG. 15 shows a bar chart of the root-mean-square-error (RMSE) for various models, in accordance with embodiments of the present disclosure. RMSE results showing the lower error with transfer learning approach with a Neural Network.

[0112] In some embodiments, the disclosed techniques may include refining a model using synthetic data. For example, the techniques may include generating synthetic data that fills in gaps or missing variables of the real data. To illustrate this, FIG. 16 shows graphs indicatingperformance of the disclosed ML backbone model that is trained using incomplete data (e.g., partial data) and complete data for different parameters described herein. In general, the graphs of FIG. 16 show that generating synthetic data to supplement real data when there are gaps in the real data may improve performance (e.g., accuracy) of the disclosed backbone model. Accordingly, generating the ML backbone model may include determining a gap exists in real data, generating synthetic data to fill in the gap (e.g., by interpolating or extrapolating), and training the ML backbone model using the combined data (e.g., the synthetic data and the real data) and / or refining the ML backbone model (e.g., further training).

[0113] As described herein, transfer learning techniques may be applied to the ML backbone model. This is generally illustrated in FIG. 17. For example, FIG. 17 shows a representation of adjusting the ML backbone model (e.g., trained using only synthetic data) to create a second model using either the feature transfer or the fine tuning approach. In either case, the graphs (e.g., scatter plots) indicate improved performance of the second model (e g., refined using transfer learning or fine tuning) because the predicted versus actual values lie along y=x as compared to the data corresponding to the ML backbone model before refining (e.g., ‘BaseReal’). A NN is represented using nodes (e.g., nodes trained using synthetic data or refined using real data). For the NN feature transfer show only the last layers of the NN being retrained while fine tuning re-trains the entire NN. In other words, the patterned circles represent the NN layers that are re-trained during transfer learning. Referring back to FIG. 13, it should be noted that utilizing these techniques may improve the computational efficiency for generating the ML backbone model or the refined ML backbone model.

[0114] FIG. 18 shows boxplots that indicate how the prediction is close to the actual ID values. Accordingly, FIG. 18 illustrates how the optimized values are significantly bettershowing the advantages of utilizing the optimizer to try to find the best value given the constraints that are able to be controlled from the fracture design parameters. Plot on the right shows the optimized against the prediction to show the advantage of the optimizer. From the predicted value we are able to modify some values and find design parameter than would lead to a higher productivity. From the plot on the right, it is apparent that the majority of the improvement comes from the middle to lower values whether higher values are not able to be optimized.

[0115] Accordingly, the disclosed techniques are directed to generating a synthetic physicsbased dataset for hydraulic fracturing or stimulation to use machine learning or predictive modeling. A processor may generate a machine-learning (ML) backbone model using the synthetic physics-based dataset and, in some instances, with a real dataset. In some instances, a processor may generate synthetic data based on gaps in the real dataset. In any case, the processor may utilize transfer learning or fine tuning approaches to further refine the ML backbone model as described with respect to FIG. 17. In turn, the model may be used to control well stimulation operations, hydraulic fracturing operations, and other oil and gas operations discussed herein.

[0116] Technical effects of the present disclosure include several benefits. For example, based on the data available for fracturing, most clients around the world have sparse data which does not yield reliable results with pure data model approaches as given in the prior art. So, using synthetic dataset could be the future of the predictive modeling within the fracturing and stimulation domain. As compared to certain conventional techniques, there is no involvement of synthetic, physics-based data here.

[0117] The specific embodiments described herein have been illustrated by way of example, and it should be understood that these embodiments may be susceptible to various modifications and alternative forms. It should be further understood that the claims are not intended to be limited to the particular forms disclosed, but rather to cover all modifications, equivalents, and alternatives falling within the spirit and scope of this disclosure.

Claims

CLAIMS1. A method, comprising: generating a synthetic physics-based dataset for hydraulic fracturing or stimulation to use machine learning or predictive modeling; generating a machine-learning (ML) backbone model using the synthetic physics-based dataset; and adjusting a hydraulic fracturing operation, a stimulation operation, or both, using the ML model.

2. The method of claim 1, where the synthetic physics-based dataset utilizes one or all of: a. reservoir rock and fluid parameters; b. geomechanical parameters; c. analyzed parameters out of diagnostic pumpings; d. fracturing fluid properties; e. proppant properties; f. perforation properties; g. fracture geometry parameters; h. fracturing treatment parameters; i. pump schedules; and j. production parameters.

3. The method of claim 1, comprising generating an inverse solution using an optimizer algorithm loop based on the synthetic physics-based dataset to enhance the fracture design in order to maximize the production performance.

4. The method of claim 1, where generating the synthetic physics-based dataset comprises building the synthetic physics-based dataset based on analytical, semi-analytical or empirical formulae and models.

5. The method of claim 1, where in the synthetic physics-based dataset is built based on a hydraulic fracturing simulator.

6. The method of claim 1, wherein generating the synthetic physics-based dataset comprises randomizing one or more parameters within a randomized range.

7. The method of claim 1, wherein adjusting the hydraulic fracturing operation, the stimulation operation, or both, using the ML model comprises: generating a pumping schedule based on the ML model; and adjusting operation of equipment by implementing the pumping schedule.

8. A method, comprising: receiving a synthetic physics-based dataset corresponding to fracturing and reservoir properties;generating a machine-learning (ML) backbone model using the synthetic physics-based dataset; receiving a real dataset corresponding to a real field; and generating a transfer learned model through feature transfer or fine tuning methods, optimizing cost functions using the real dataset and the synthetic physics-based dataset.

9. The method of claim 8, wherein generating the ML model comprises splitting predictors and outputs corresponding to the synthetic physics-based dataset.

10. The method of claim 8, comprising generating a fracturing treatment and production performance outputs for a well based on either the ML backbone model or the ML transfer learned model, where the transfer learning is performed through either feature training or fine tuning techniques11. The method of claim 10, wherein generating the fracturing treatment output comprises: receiving one or more of input reservoir properties, input calibration injection parameters, input fracturing treatment properties, input proppant properties, and input perforation properties; determining one or more of a pad ratio, a proppant mass (MproP), a maximum proppant concentration (PPA), a treatment pump schedule, and a dimensionless productivity index (JD) using the ML transfer learned model based on the one or more of the input reservoir properties, the input calibration injection parameters, the input fracturing treatment properties, the input proppant properties, and the input perforation properties;generating the fracturing treatment output using the one or more of the pad ratio, the proppant mass (Mprop), the maximum proppant concentration (PPA), the treatment pump schedule, and the dimensionless productivity index (JD).

12. The method of claim 8, comprising combining a real fracturing dataset with the synthetic physics-based dataset using a transfer learning approach.

13. The method of claim 8, wherein generating the transfer learned model through feature transfer comprises refining the ML backbone model.

14. The method of claim 8, comprising: utilizing classical techniques and correlations to generate one or more dependent properties based on randomized one or more parameters; and generating the ML backbone model using the one or more dependent properties.

15. A system, comprising: one or more memory storing a physics-based model to predict unseating of diversion material based on drag forces, pressure forces, buoyancy forces, cake cohesion, chemical degradation of particulates, or a combination thereof. a control system comprising one or more processors, wherein the control system is configured to: receive a synthetic dataset corresponding to fracturing and reservoir properties; generate a machine-learning (ML) backbone model using the synthetic dataset;receive a real dataset corresponding to a real field; generate an ML transfer learned model using feature training or fine tuning by optimizing cost functions using the real dataset and the synthetic dataset; and adjust a hydraulic fracturing operation, a stimulation operation, or both, using the ML model.

16. The system of claim 15, wherein the control system is configured to adjust the hydraulic fracturing operation, the stimulation operation, or both, using the ML model by: generating a pumping schedule based on the ML model; and adjusting operation of equipment by implementing the pumping schedule.

17. The system of claim 15, wherein the control system is configured to generate a fracturing treatment output based on the ML model and adjust the hydraulic fracturing operation, the stimulation operation, or both using the fracturing treatment output.

18. The system of claim 17, wherein generating the fracturing treatment output comprises: receiving one or more of input reservoir properties, input calibration injection parameters, input fracturing treatment properties, input proppant properties, and input perforation properties; determining one or more of a pad ratio, a proppant mass (Mprop), a maximum proppant concentration (PPA), a treatment pump schedule, and a dimensionless productivity index (JD) using the ML backbone model based on the one or more of the input reservoir properties, theinput calibration injection parameters, the input fracturing treatment properties, the input proppant properties, and the input perforation properties; generating the fracturing treatment output using the one or more of the pad ratio, the proppant mass (MprOp), the maximum proppant concentration (PPA), the treatment pump schedule, and the dimensionless productivity index (JD).

19. The system of claim 15, wherein the control system is configured to generate the synthetic dataset by randomizing one or more parameters within a randomized range.

20. The system of claim 15, wherein the control system is configured to adjust the hydraulic fracturing operation, the stimulation operation, or both, using the ML model by: obtaining a bridging condition for fracture width; determining a maximum proppant concentration based on the bridging condition; and adjusting the hydraulic fracturing operation using the maximum proppant concentration.

Images