A COMPUTER-READABLE METHOD FOR CONTROLLING SURFACE SYSTEMS FOR HYDROCARBON RESERVOIRS, SYSTEM, AND STORAGE MEDIUM

Abstract

AUTONOMOUS SURFACE SYSTEM. The surface system control methods described in this document include acquiring data during the operation of a fluid-handling surface system using subsystems; generating a parameter value for one of the subsystems using at least some of the data and a model that models the subsystem using historical data that includes operational tuning data for operational adjustments in response to conditions; and automatically controlling one of the subsystems using the parameter value.

Description

1 / 57 A COMPUTER-READABLE METHOD FOR CONTROLLING SURFACE SYSTEMS FOR HYDROCARBON RESERVOIRS, SYSTEM, AND STORAGE MEDIUM RELATED REQUEST

[0001] This application claims priority and the benefit of a U.S. Provisional Application with Serial No. 62 / 982766, filed February 28, 2020, which is incorporated by reference herein. FUNDAMENTALS Field

[0002] The embodiments described in this document generally refer to surface systems for hydrocarbon reservoirs. Specifically, the embodiments described in this document refer to the control of such surface systems. Description of the Related Technique

[0003] The global oil and gas industry is trending toward improving environmental safety and compliance throughout the various phases of a well lifecycle. Various stages include the use of equipment, which may include equipment that is manually operated by one or more crew members, where equipment placement and / or equipment operation may present risks. SUMMARY

[0004] The embodiments described in this document provide methods for controlling surface systems that may include acquiring data during the operation of a fluid-handling surface system using subsystems; generating a parameter value for one of the subsystems using at least a portion of the data and a model that models the subsystem using historical data that includes operational tuning data for condition-responsive operational adjustments; and automatically controlling one of the subsystems using the parameter value. Several other examples of methods, systems, computer program products, etc., are also described in this document. BRIEF DESCRIPTION OF THE DRAWINGS Petition 870260069218, dated 07 / 13 / 2026, page 15 / 136 2 / 57

[0005] In order that the above-mentioned features of the present disclosure may be understood in detail, a more particular description of the disclosure, briefly summarized above, may be made by reference to the embodiments, some of which are illustrated in the accompanying drawings. However, it will be noted that the accompanying drawings illustrate only exemplary embodiments and, therefore, will not be considered limiting of its scope and may admit other equally effective embodiments.

[0006] Fig. 1 is a series of example environment diagrams and an example of a surface system.

[0007] Fig. 2 is a diagram of an example of a surface system.

[0008] Fig. 3 is a diagram of an example of a system.

[0009] Fig. 4 is a diagram of an example of a control system.

[0010] Fig. 5 is a diagram of an example of a flow meter.

[0011] Fig. 6 is a diagram of an example of a system.

[0012] Fig. 7 is a diagram of an example of a system.

[0013] Fig. 8 is a diagram of an example of a model.

[0014] Fig. 9 is a diagram of an example of a machine learning model.

[0015] Fig. 10 is a diagram of an example of a computational structure.

[0016] Fig. 11 is a diagram of an example of a method.

[0017] Fig. 12 is a diagram of an example of a method and an example of a system.

[0018] Fig. 13 is a diagram of an example of a computer system.

[0019] Fig. 14 is an example component diagram of a system and a networked system.

[0020] To facilitate understanding, identical reference numerals have been used, where possible, to designate identical elements that are common to the figures. It is contemplated that elements and characteristics of a modality may be incorporated. Petition 870260069218, dated 07 / 13 / 2026, page 16 / 136 3 / 57 in a beneficial way in other modalities without additional recitation. DETAILED DESCRIPTION

[0021] The following description includes the best currently contemplated method for practicing the implementations described. This description should not be taken in a limiting sense, but is made only for the purpose of describing the general principles of the implementations. The scope of the implementations described should be determined with reference to the claims made.

[0022] Fig. 1 shows examples of environments 101, including a marine environment 102 and a terrestrial environment 104 where the marine environment 102 includes various equipment and where the terrestrial environment 104 includes various equipment. As shown, each of the environments 101 may include one or more wellheads 106 (e.g., wellhead equipment). A wellhead may be a surface termination of a wellbore that may include a system of assorted spools, valves, and adapters that, for example, may provide pressure control of a production well. A wellhead may be on a land surface, a subsea surface (e.g., an ocean floor, etc.), etc. As an example, a wellhead may include one or more valves, such as, for example, one or more choke valves. A choke valve may be located on or near a Christmas tree that is used to control the fluid production of a well.For example, opening or closing a variable valve can influence the rate and pressure at which production fluids progress through a pipeline, process facilities, etc. As an example, an adjustable choke can be operationally coupled to an automated control system to allow one or more production parameters of one or more individual wells to be controlled. As an example, multi-wellhead conduits can be joined into one or more manifolds so that fluid from multiple wells can flow in a common conduit.

[0023] As shown, in various environments, during one or more phases of development, production, etc., surface equipment may be present that is in fluid communication with a well, a completed well, etc. Such surface equipment (e.g., a surface system) may be in fluid communication Petition 870260069218, dated 07 / 13 / 2026, page 17 / 136 4 / 57 for fluid injection and / or fluid production purposes. For example, fluid injection may include hydraulic fracturing fluid injection to generate fractures in a reservoir to increase the production of hydrocarbon-containing fluids from the reservoir, treatment fluid injection, such as a fluid for stimulation purposes, etc. As for fluid production, surface equipment may include various types of conduits, valves, meters, separators, etc. As an example, a surface system may include equipment that may be independent in its operation and / or control. For example, a subsystem may be mounted on casters with a supplied control unit. In such an example, a comprehensive control system may be operationally coupled to the control unit.When a surface system includes multiple subsystems, each may include its own controller unit and / or interface that can be operationally coupled to an overarching controller system.

[0024] In several cases, however, a comprehensive approach to the control system may make supervisory control decisions that can impact a subsystem where the subsystem may be left to its own devices as to how it handles or responds to a supervisory control decision. For example, consider an approach that aims to adequately control one or more setpoints (e.g., pressure, level, etc.) and that may take higher-level actions as appropriate, such as regulating flow to remain within a pressure / flow range of the equipment.

[0025] In several examples, a self-contained surface system is described in relation to the surface equipment associated with well testing, noting that, as mentioned, one or more other types of surface systems may be similarly instrumented to be a self-contained surface system for one or more purposes.

[0026] Referring again to Fig. 1, at various times a well may be tested using a process referred to as well testing. Well testing may include one or more of a variety of well testing operations. In several cases, fluid may flow from a well or wells to the surface where the fluid is subjected to one or more well testing operations and generates scrap (e.g., residual fluid), which must be handled appropriately, for example, according to circumstances, regulations, etc. For example, consider loading residual fluid onto a tanker to Petition 870260069218, dated 07 / 13 / 2026, page 18 / 136 5 / 57 transport to a facility that can dispose of the residual fluid. Another way to handle residual fluid can be through combustion, which can be referred to as burning. As an example, burning can be part of a well testing process if the burning is to handle residual fluid and / or to analyze one or more aspects of how one or more residual fluids burn. Regarding the latter, burning can optionally provide data regarding one or more characteristics of the well fluid (e.g., a component thereof, etc.).

[0027] As an example, well testing can be carried out during one or more phases, such as during exploration and evaluation, where hydrocarbon production is tested using a temporary production facility that can provide fluid sampling, flow analysis, and pressure information generation, for example, to help characterize a reservoir. Several decisions can be based on well testing, such as decisions regarding production methods, facilities, and potential well productivity improvements.

[0028] Regarding example environments 101 in Fig. 1, well testing can be performed, for example, using the equipment shown in marine environment 102 and / or using the equipment shown in terrestrial environment 104. As an example, an environment may be in exploration, development, evaluation, etc., where such environment includes at least one well where well fluid can be produced (e.g., via natural pressure, via fracturing, via artificial lift, via pumping, via flooding, etc.). In such an environment, various types of equipment may be on site, which can be operationally coupled to the well testing equipment.

[0029] Fig. 1 shows an example of a system 110 (e.g., a surface system) that can be operatively coupled to one or more conduits that can transport well fluid, for example, from one or more wellheads. As shown, the system 110 may include a computing system 111 (CS), which may include one or more processors 112, memory 114 accessible to at least one of the one or more processors 112, instructions 116 that can be stored in memory 114 and executable by at least one of the one or more processors 112, and one or more interfaces 118 (e.g., wired, wireless, etc.) that can be used, by Petition 870260069218, dated 07 / 13 / 2026, p. 19 / 136 6 / 57 example, for one or more types of communication with one or more of the different subsystems and / or pieces of equipment of the surface system. In the example of Fig. 1, system 110 is shown as including various communication symbols, which may be for transmission and / or reception of information (e.g., data, commands, etc.), for example, to and / or from computer system 111. As an example, computer system 111 may be a controller that can issue control instructions to one or more pieces of equipment in an environment such as, for example, the marine environment 102 and / or the terrestrial environment 104. As an example, computer system 111 may be local, may be remote, or may be distributed (e.g., partly local and partly remote, multiple locations and / or remote locations, etc.).

[0030] Referring again to wellhead 106, it may include various types of wellhead equipment, such as, for example, casing and tubing heads, a production tree, a blowout preventer, etc. The fluid produced from a well may be routed through wellhead 106 and to system 110, which may be configured with various features for well testing operations.

[0031] In the example in Fig. 1, system 110 is shown to include several segments, which can be operationally categorized. For example, consider a well control segment 120, a separation segment 122, a fluid management segment 124, and a flare segment 126. In such an example, one or more of the several segments may correspond to a subsystem or subsystems. For example, consider the separation segment 122 corresponding to a separation subsystem.

[0032] As shown in the example in Fig. 1, the well control segment 120 is an assembly of several components, such as a manifold 130, a throttling manifold 132, a manifold 134, a heat exchanger 136, and a meter 138; the separation segment 122 includes a separator 142; the fluid management segment 124 is an assembly of several components, such as pump manifolds and pumps 144, a tank manifold 146-1, a tank manifold 146-2, a tank 148-1, and a tank 148-2; and the flare segment 126 includes a flare 152 and one or more chambers 154. In several examples, a manifold may be an arrangement of pipes and valves. Petition 870260069218, dated 07 / 13 / 2026, page 20 / 136 7 / 57 for controlling fluid circulation. A tank manifold can allow control of fluid in and / or out of a tank, while a pump manifold can allow control of fluid in and / or out of pumps.

[0033] As mentioned in the example in Fig. 1, the 110 system includes several features for one or more aspects of well testing operations; noting that the 110 system may include minor features, more features, alternative features, etc. In particular, each segment may include one or more sensors associated with one or more pieces of equipment and / or locations in the segment. The sensors may detect parameters (e.g., variables, etc.) such as temperature, pressure, flow, or the state of a piece of equipment (e.g., the state of a valve). One or more other types of sensors may also be used as part of the system. For example, consider one or more of a gas specific gravity meter, a water cutoff meter, a gas / oil ratio sensor, a carbon dioxide sensor, a hydrogen sulfide sensor, or a shrinkage measuring device.Various features may be located upstream and / or downstream of a separator segment or a separator itself.

[0034] With regard to the fluid flow from a well or wells, such fluid may be received by the well control segment 120 and then routed through one or more conduits to the separation segment 122. In the example of Fig. 1, the heat exchanger 136 may be provided as a steam heat exchanger and the meter 138 to measure the fluid flow through the well control segment 120.

[0035] As mentioned, well control segment 120 can transport fluid received from one or more wells to separator 142. As an example, separator 142 can be a horizontal separator or a vertical separator and can be a two-phase separator (e.g., to separate gas and liquids) or a three-phase separator (e.g., to separate gas, oil, and water). A separator may include various features to facilitate the separation of components of the incoming fluid (e.g., diffusers, mist extractors, vanes, deflectors, precipitators, etc.).

[0036] As an example, the fluid can be a single-phase or multi-phase fluid where “phase” can refer to an immiscible component (for example, consider two or Petition 870260069218, dated 07 / 13 / 2026, page 21 / 136 8 / 57 more oil, water and gas for a multiphase fluid).

[0037] As an example, separator 142 can be used to substantially separate multiphase fluid into its oil, gas, and water phases, as appropriate and present, where each phase emerging from separator 142 can be referred to as a separated fluid. Such separated fluids can be routed away from separator 142 to fluid management segment 124. In several cases, the separated fluids may not be entirely homogeneous. For example, the separated gas exiting separator 142 may include some residual amount of water or oil, and the separated water exiting separator 142 may include some amount of entrained oil or gas. Similarly, the separated oil exiting separator 142 may include some amount of entrained water or gas.

[0038] As shown in the example in Fig. 1, a system may include one or more manifolds, where depending on the number of wells (e.g., 1, 2, 3, ..., N), types of equipment, etc., a single manifold may be sufficient or there may be more than one manifold. In the example in Fig. 1, the fluid management segment 124 may include flow control equipment, such as one or more manifolds and one or more pumps (usually represented by block 144) to receive fluids from separator 142 and transport the fluids to other destinations, optionally along with one or more additional manifolds 146-1 and 146-2, for example, to direct fluid to and from fluid tanks 148-1 and 148-2. As explained, the number of manifolds and tanks may vary according to several factors.For example, in one embodiment, the 124 fluid management segment may include a single manifold and a single tank, while in other embodiments the 124 fluid management segment may include more than two manifolds and more than two tanks.

[0039] As for the manifolds and pumps 144, they may include a variety of manifolds and pumps, such as a gas manifold, an oil manifold, an oil transfer pump, a water manifold, and a water transfer pump. In at least some embodiments, the manifolds and pumps 144 may be used to direct fluids received from the separator 142 to one or more of the fluid tanks 148-1 and 148-2 via one or more of the additional manifolds 146-1 and 146-2 and to direct fluids Petition 870260069218, dated 07 / 13 / 2026, page 22 / 136 9 / 57 between tanks 148-1 and 148-2. As an example, the manifolds and pumps 144 may include features for directing fluids received from separator 142 directly to one or more burners 152 for burning gas and oil (e.g., bypassing tanks 148-1 and 148-2) or for directing fluids from one or more of tanks 148-1 and 148-2 to one or more burners 152.

[0040] As noted above, the components of system 110 may vary between different applications. As an example, the equipment within each functional group of system 110 may also vary. For example, heat exchanger 136 may be supplied as part of separation segment 122, rather than well control segment 120.

[0041] In certain embodiments, system 110 may be a surface well testing system that can be remotely monitored and controlled. Remote monitoring may be effected with sensors installed on various components. In some cases, a monitoring system (e.g., sensors, communication systems, and human-machine interfaces) may allow the monitoring of one or more of segments 120, 122, 124, and 126. As shown in the example in Fig. 1, one or more cameras 154 may be used to monitor one or more firing operations of one or more burners 152, which may aim to facilitate the control of such one or more firing operations at least in part through the analysis of image data acquired by at least one of the one or more cameras 154. As an example, one or more cameras may be used for temperature monitoring.For example, consider an infrared camera that can use infrared wavelength emissions (e.g., consider approximately 1 μm to approximately 14 μm) to determine temperature, where temperature can be used for process control, safety, etc.

[0042] Fig. 2 shows an example of a 250 system, which can be referred to as a surface well test system. The 250 system may include several features of the 110 system in Fig. 1.

[0043] In Fig. 2, a multiphase fluid (represented here by arrow 252) enters a flow header 254 and is routed to a separator 270 through a Petition 870260069218, dated 07 / 13 / 2026, page 23 / 136 10 / 57 surface safety valve 256, a steam heat exchanger 260, a throttling manifold 262, a flow meter 264 and an additional manifold 266. In the example of Fig. 2, the system 250 includes a chemical injection pump 258 for injecting chemicals into the multiphase fluid flowing towards the separator 270, as may be desired.

[0044] In the embodiment shown in Fig. 2, separator 270 is a three-phase separator that generally separates multiphase fluid 252 into gas, oil, and water components. The separated gas is routed downstream of separator 270 through a gas manifold 274 to either burner 276-1 or 276-2 for combustion gas and oil. The gas manifold 274 includes valves that can be actuated to control the gas flow from the gas manifold 274 to either burner 276-1 or 276-2. Although shown side-by-side in Fig. 2 for clarity, burners 276-1 and 276-2 can be positioned separately from each other, such as on opposite sides of a probe, etc.

[0045] As shown, the oil separated from separator 270 can be routed downstream to an oil collector 280. The valves of oil collector 280 can be operated to allow the flow of oil to either of burners 2761 and 276-2 or either of tanks 282 and 284. Tanks 282 and 284 can be of a suitable shape, but are represented in Fig. 2 as vertical surge tanks, each having two fluid compartments. Such an approach allows each of tanks 282 and 284 to simultaneously hold different fluids, such as water in one compartment and oil in the other compartment. An oil transfer pump 286 can be operated to pump oil through the well test system 250 downstream of separator 270. The water separated from separator 270 can be similarly routed to a water collector 290.Like the oil collector 280, the water collector 290 includes valves that can be opened or closed to allow water to flow to one of the tanks 282 and 284, or to a water treatment and disposal unit 294. A water transfer pump 292 can be used to pump water through the system.

[0046] A well testing area in which the 250 well testing system (or Petition 870260069218, dated 07 / 13 / 2026, page 24 / 136 11 / 57 other embodiments of a well testing system) where it is installed may be classified as a hazardous area. In some embodiments, the well testing area is classified as a Zone 1 hazardous area, according to the International Electrotechnical Commission (IEC) standard 60079-10-1:2015.

[0047] In the example of Fig. 2, a cabin 296 at a well site may include various types of equipment for acquiring data from the well test system 250. This acquired data may be used to monitor and control the well test system 250. In at least some cases, the cabin 296 may be separated from the well test area by having the well test system 250 in a non-hazardous area. This is represented by the dashed line 298 in Fig. 2, which generally serves as a demarcation between the hazardous area with the well test system 250 and the non-hazardous area of ​​the cabin 296.

[0048] The equipment of a well testing system can be monitored during a well testing process to verify proper operation and facilitate process control. Such monitoring may include various measurements by appropriate sensors during a well test, examples of which may include throttling manifold temperature and pressures (upstream and downstream), heat exchanger temperature and pressure, separator temperature and pressures (static and differential), separator oil flow and volume, separator water flow and volume, and fluid levels in the tanks of a system.

[0049] As an example, a system can be configured for local and / or remote processing of information, control, etc. For example, consider a mobile computing device, such as a tablet computing device, that can be operationally coupled to remote computing resources via a wired network, a wireless network, etc. In such an example, the remote computing resources may be or include a multi-cloud management platform (MCMP, for example, an IBM MCMP, etc.; International Business Machines Corporation, Armonk, New York). In such an example, a mobile computing device may include hardware suitable for running a browser application or other type of application suitable for rendering graphical user interfaces to a display, which may be a screen display. Petition 870260069218, dated 07 / 13 / 2026, page 25 / 136 12 / 57 touch-sensitive. For example, consider a browser application running on a mobile computing device that a user can interact with a MCMP for one or more purposes. In such an approach, the mobile computing device can provide interactions for one or more of the following: equipment maintenance, equipment sensor data, equipment control (e.g., setpoints, etc.), etc. In such an approach, a user can evaluate equipment using the mobile computing device, which can provide flexibility to the user regarding the user's location, which can be, for example, remote from an equipment location. Using the mobile computing device, a user can "check in" various types of equipment that are at a location on a daily basis or on a less frequent basis and / or on a more frequent basis.

[0050] Fig. 3 shows an example of a 300 system that includes a hazardous area 301 and a non-hazardous area 303; noting that various types of computing equipment, network equipment, etc., can be positioned or repositioned in one or more of the 301 and 303 areas. As mentioned, a cabin can be included in an area, which can help protect one or more people, equipment, etc., from one or more hazards, or, for example, a cabin can be in an area characterized as non-hazardous.

[0051] In the example in Fig. 3, the 300 system is illustrated as an architecture with various types of equipment. For example, a 310 environment is shown as including equipment that can perform various actions related to well operations, such as well testing.

[0052] When such a system does not include several features of system 300, one or more operators may be present for one or more manual tasks as for operations in environment 310. Such tasks may be referred to as jobs, which may be designated using the French word “métier,” meaning work, for example, the work of testing a well. In performing such tasks (e.g., jobs), an operator may have knowledge and experience of how the equipment behaves under certain conditions, how the fluid behaves under certain conditions, how combustion behaves under certain conditions, etc. Such an operator may be instructed or understand how Petition 870260069218, dated 07 / 13 / 2026, p. 26 / 136 13 / 57 Taking one or more actions in the 310 environment, which may be for the optimization of one or more processes and / or for risk reduction, for example, in an emergency situation. As a system may include several subsystems, coordinated action may be required to properly optimize and / or reduce risk. However, when coordinated action is through a crew, there may be considerable demands placed on crew members, particularly with regard to timing, adjustments, communications, etc. For example, an action taken by a first operator in a first subsystem may affect how fluid flows to a second subsystem, which may be managed by a second operator. If the second operator does not anticipate the impact, the second operator may view the changes as an emergency and request a system-wide shutdown, or the second operator may make one or more changes that cascade to one or more other subsystems.

[0053] In a system, comprehensive control, which may be referred to as supervisory control, and subsystem control may optionally be implemented with one or more independent safety systems. In such an example, one or more digital twins with métier (DTMs) may be implemented, which may operate at least in part at a subsystem level. A DTM is a model that virtually replicates one or more pieces of equipment at a well site. Such a DTM may be used to run simulations and be trained to determine optimal equipment and / or system behavior based on current detected parameters. For example, consider a separator DTM that has been trained (e.g., through machine learning, etc.) to possess the knowledge and experience of an operator versed in separator operation. In such an example, the DTM may be located and operate in a manner that is based on knowledge, experience, and data.For example, sensor data can be acquired for the separator and fed into a model that can produce one or more parameter values ​​that can be used to control the separator's operation. In such an example, the separator can be controlled as an autonomous surface system (e.g., an autonomous surface subsystem). Such a DTM approach can be robust and capable of handling events such as, for example, a shutdown at a wellhead, optionally without receiving. Petition 870260069218, dated 07 / 13 / 2026, page 27 / 136 14 / 57 of a communication of this event. For example, the DTM can be robust insofar as it can respond to locally acquired data and understand which parameter values ​​will result in optimal separator operation, whether for well testing purposes or other operation(s) (e.g., shutdown, startup, etc.).

[0054] In the example in Fig. 3, environment 310 includes flow control equipment 311 (e.g., capable of performing a shutdown and / or other process action), which may be in fluid communication with one or more wells, for example, to supply well fluid to one or more pieces of equipment in environment 310. As shown, environment 310 may include various data acquisition equipment 312, one or more assets 314 and 316 (e.g., subsystems, etc.) and a safety system 320 operatively coupled to the flow control equipment 311. As an example, when a condition, or conditions, arise that warrant safety action, the safety system may issue a control instruction to shut off the fluid flow using the flow control equipment 311 (e.g., a throttling valve, etc.).

[0055] In the example in Fig. 3, the 310 environment may include one or more types of controllers, which may be operationally coupled to various equipment. Such controllers (e.g., control units, control systems, etc.) may be manufactured and / or otherwise protected for operation in the 310 environment.

[0056] As shown, the 300 system may include a 330 controller, which may be a control system that is operationally coupled to the equipment in the 310 environment through one or more communication technologies. For example, consider wired and / or wireless technologies, which may use one or more types of communication protocols. For example, consider EtherNet / IP where “IP” is an abbreviation for “Industrial Protocol,” which is an industrial network protocol that adapts the Common Industrial Protocol (CIP) to Ethernet. As an example, the 330 controller may be a supervisory-level controller.

[0057] For example, in system 300, an operator may be present in environment 310 or in the vicinity of environment 310 and use one or more computing devices 350, which may be operationally coupled to equipment in the area. Petition 870260069218, dated 07 / 13 / 2026, p. 28 / 136 15 / 57 dangerous 301 and / or in the non-dangerous area 303 (e.g., directly, indirectly, etc.) through one or more interfaces. For example, consider a mobile device that can communicate with the controller 330 and / or the safety system 320, which can be via direct and / or indirect communication (e.g., wired and / or wireless).

[0058] Fig. 3 also shows various communications that can use one or more technologies, such as, for example, one or more of HTTPS, Remote Desktop Protocol (RDP), an OPC Unified Architecture (OPC-UA), etc.

[0059] The Secure Hypertext Transfer Protocol (HTTPS) is a Hypertext Transfer Protocol (HTTP) and can be used for secure communication over a computer network. In HTTPS, the communication protocol can be encrypted using a security technology such as Transport Layer Security (TLS) or Secure Sockets Layer (SSL).

[0060] Remote Desktop Protocol (RDP) is a proprietary protocol developed by Microsoft Corporation (Redmond, Washington) that can provide a graphical interface for connecting to another computer over a network connection. RDP involves the use of RDP client software and RDP server software (e.g., also consider HTTPS, etc.). Clients exist for various operating systems (OSs) (e.g., WINDOWS, LINUX, UNIX, macOS, iOS, ANDROID, etc.). RDP servers can be embedded in an operating system (e.g., WINDOWS operating system, UNIX operating system, etc.), where a server can listen on a TCP port and a UDP port. RDP is an extension of the ITU-T T.128 Application Sharing Protocol.

[0061] OPC-UA is a machine-to-machine communication protocol for industrial automation developed by the OPC Foundation. It can provide communication between industrial equipment and systems for data collection and control, is cross-platform, provides a service-oriented architecture (SOA), and provides various security measures.

[0062] OPC-UA can use an integral information model to model data in an OPC-UA namespace for SOA. OPC-UA supports protocols Petition 870260069218, dated 07 / 13 / 2026, p. 29 / 136 16 / 57 as a binary protocol (e.g., server opc.tcp: / / ) and an HTTP protocol (e.g., server http: / / ) for web service. OPC-UA can operate transparently to an application programming interface (API). As an example, EtherNet / IP can be used, optionally according to one or more API specifications.

[0063] A binary protocol can offer less overhead and require fewer resources (e.g., no XML parser, SOAP, and HTTP), which can facilitate operations for embedded devices. A binary protocol can offer interoperability and use a single, arbitrarily chosen TCP port to facilitate tunneling communication or easy enabling through a firewall.

[0064] As an example, the web service protocol (SOAP) can be used and supported from available tools (e.g., JAVA or .NETWORK environments) and can be firewall compatible (e.g., using HTTP(S) ports).

[0065] In the example of Fig. 3, the non-hazardous area 303 is shown as including several systems 360, 370 and 380, together with a digital dual métier (DTM) 390 (e.g., or a plurality of DTMs, etc.). The DTM 390 can be constructed using one or more systems and then deployed for local use relative to the well site equipment in the environment 310. For example, the DTM 390 can be a model, which can be, for example, an algorithm (e.g., manually coded decision tree, etc.) or a trained machine learning model (trained ML model) that can be deployed on one or more controllers, which may be in the hazardous area 301 and / or optionally in a cabin, etc., which may be a “safe” area within the hazardous area 301 or in an area adjacent to the hazardous area 301.

[0066] As an example, the DTM 390 can be deployed on the controller 330, which, as mentioned, can be a primary controller for the environment 310 (e.g., a supervisory level controller). As explained, for safety reasons, the safety system 320 can operate where one or more conditions arise that could increase risk. For example, the safety system 320 can respond to the controller 330 failing (e.g., power loss, operational error, etc.) and / or it can Petition 870260069218, dated 07 / 13 / 2026, page 30 / 136 17 / 57 responding to the controller being unable to sufficiently control a condition or conditions, which may include one or more types of environmental trend conditions 310. As for operations, the safety system 320 may cause interruption of fluid flow through flow control equipment 311 (e.g., a throttling valve, etc.).

[0067] As mentioned, the DTM 390 can be representative of a subsystem within the 310 environment, where it can provide a self-contained surface subsystem. As mentioned, a separator can be modeled as a digital twin, where the operator's knowledge and experience embedded in tasks, jobs, etc., to be performed by the operator are also modeled so that the digital twin is a métier digital twin (DTM). As mentioned, the 310 environment can include multiple subsystems, and one or more DTMs can each be representative of one or more of the subsystems.

[0068] Regarding various components, systems, etc., in the non-hazardous area 303, as shown in Fig. 3, the 360 ​​system may include a 362 component for a well site structure and one or more web applications, which may be operationally coupled to a 364 component for a data structure that may include or be operationally coupled to one or more databases (DBs). As shown, the 360 ​​system may include a 366 component for a web server that may operate according to a supervisory control and data acquisition (SCADA) architecture, which may provide interactions between computing devices, controllers, networked data communication equipment, graphical user interfaces (GUIs), etc.A SCADA architecture can provide high-level process supervision management, as well as include other peripheral devices such as programmable logic controllers (PLCs) and discrete proportional-integral (PID) controllers to interface with the equipment.

[0069] In the example in Fig. 3, the 360 ​​system may be communicating with the 330 controller (e.g., OPC-UA, etc.) and through one or more of the 350 operator devices (e.g., RDP, HTTPS, etc.). The 360 ​​system may use one or more virtualization technologies, such as, for example, virtual machines (VMs) and / or Petition 870260069218, dated 07 / 13 / 2026, page 31 / 136 18 / 57 Containerization. As an example, a system may include hardware virtual machines and / or process virtual machines.

[0070] For example, a VM can run a complete operating system, including its own kernel. For example, a container can be an isolated and lightweight silo for running an application on a host operating system (host OS). For example, a container can be built on top of a host OS kernel and include applications and, for example, some lightweight OS APIs and services that can run in user mode.

[0071] The 360 ​​system can be implemented using various resources, which may include cloud-based resources. For example, the 360 ​​system can be partially implemented using cloud-based resources (e.g., servers from a server farm, data storage devices from a server farm, etc.). For example, the 360 ​​system can be accessible through one or more protocols (e.g., wired or wireless) so that remote interactions can occur (e.g., for remote management, etc.), which can be through a cloud environment.

[0072] As for component 364, it may include various features of the Rockwell Automation suite (e.g., FactoryTalk suite, etc., Rockwell Automation, Milwaukee, Wisconsin). Such features may be suitable for interactions with a controller system, a control unit, etc., which may be a Rockwell Automation controller system, control unit, etc. (e.g., consider one or more Allen-Bradley products, etc.).

[0073] As an example, component 364 can provide data organization at the equipment and / or enterprise level. Component 364 can include historian capabilities to collect time series data for various calculations, estimates, and statistical processes. Component 364 can provide trend reports and analyses.

[0074] As an example, component 364 can provide predictions, such as, for example, anomaly predictions, equipment degradation predictions, etc. As an example, component 364 can include an embedded analysis feature, which Petition 870260069218, dated 07 / 13 / 2026, page 32 / 136 19 / 57 can provide analytics for use in training a machine learning model, operating a trained machine learning model, etc.

[0075] In the example in Fig. 3, the 370 system can include various resources for media, applications, and dockers, which can be operationally coupled to a component labeled as a message bus / message queue, which can be a message broker resource (e.g., message-oriented middleware) that can implement the Advanced Message Queuing Protocol (AMQP) and be extended with a plug-in architecture to support Text-Oriented Message Transmission Protocol (STOMP), MQTT, and other protocols. Such a component can be operable using a LINUX operating system environment, which can be implemented using a component such as a multi-cloud management platform (MCMP, International Business Machines Corporation, Armonk, New York), which can utilize one or more servers. MCMP can be operable using one or more cloud environments and be accessible to the browser through one or more browser applications.

[0076] In the example in Fig. 3, the 380 system can include several features of the 360 ​​system, such as, for example, a data structure, a data logger (e.g., a historian, etc.), a web server (e.g., SCADA, etc.), and a well site structure. As an example, the 380 system can be a cloud-based system that can provide the construction of one or more 390 DTMs, which can then be deployed to a specific well site or sites. As an example, a particular well site might include an instance of the 380 system appropriately sized for that particular well site. As an example, the 360 ​​system might include features for selecting the 390 DTM, constructing the 390 DTM, adapting the 390 DTM, deploying the 390 DTM, operating the 390 DTM, etc.

[0077] In the example in Fig. 3, as explained, the DTM 390 can be used to control one or more pieces of equipment in the 310 environment. As mentioned, the DTM 390 can be a relatively lightweight object that can be implemented using an operating system of one or more pieces of equipment, a controller unit, a controller system, etc. As explained, the 310 environment can include Petition 870260069218, dated 07 / 13 / 2026, page 33 / 136 20 / 57 The 320 safety system is used to ensure that safety guidelines are implemented where, for example, a problem may arise where control via the 390 DTM and / or one or more other control mechanisms may be inadequate. As explained, a subsystem in the 310 environment may include a specialized DTM that can be robust and provide autonomous operation of the subsystem where an event may occur, such as a shutdown by the 320 safety system. In such an example, safety within the 310 environment can be enhanced, since a DTM subsystem can provide control of the subsystem in a way that reduces the risk to the subsystem, to one or more operators (e.g., if present or present later), to the environment (e.g., spill, flame, etc.), etc.

[0078] As an example, the 300 system may include logic, for example, at a PLC level, which may be sufficient to manage a safe shutdown. For example, consider a scenario where a network connection is lost or compromised. In such an example, DTM control may be confounded so that logic at a PLC level can address such a scenario, particularly where an emergency shutdown (ESD) system operates to shut off the flow from a well. In such an example, consider a separator as a subsystem that has been in fluid communication with the well where the separator itself can be adequately controlled, for example, to have a controlled shutdown aimed at reducing one or more hazards, such as a spill hazard, an overpressure hazard, etc.

[0079] Fig. 4 shows an example of a 400 controller system that includes several units. As an example, the 400 controller system may include several units that can be mounted in a way where the units can be operationally coupled for one or more purposes. As shown in Fig. 4, the 400 controller system includes a controller unit 401, an AC input unit 402, a communication unit 403, an AC output unit 404, a DC input unit 405, a DC output unit 406, and one or more other units 407.

[0080] As an example, a DC input unit may allow the connection of PNP (supply) and / or NPN (drain) transistor type devices (e.g., a sensor, a switch, etc.). As an example, an AC input unit Petition 870260069218, dated 07 / 13 / 2026, p. 34 / 136 21 / 57 can handle unbiased AC voltage where, for example, the AC voltage is being switched via a limit switch or other type of switch. AC input units tend to be less common than DC input units, as many sensors may have transistor output(s). For example, a sensor may be operating on a DC voltage and provide a DC output that can be received via a DC input unit.

[0081] The controller system 400 may be suitable for use as the controller 330 of the system 300 in Fig. 3 and / or for use as a subsystem controller (see, for example, assets 314, 316, etc.).

[0082] As an example, the 400 controller system may include one or more types of circuits, features, etc., of a controller system (e.g., compact logic, PLC, etc.). As an example, a controller system may include a controller unit, a communication unit, a power supply unit, one or more discrete input units, one or more removable terminal blocks for a discrete input unit, a discrete output unit, one or more removable terminal blocks for a discrete output unit, an analog input unit, executable instructions stored in memory, one or more redundant units (e.g., for redundant control, redundant power, redundant communication, etc.), etc.

[0083] As an example, a controller unit, a controller system, etc., may be or include one or more programmable logic controller (PLC) units. As an example, a controller system may be configured with particular units for dedicated use, for example, as a safety controller that may require one or more types of safety-related actions. Such a controller may be independent of one or more other controllers, so that when a primary controller fails, the safety controller may be independent and take appropriate action. As for a failure of a primary controller, such failure may be for one or more reasons, which may include, for example, failure of the controller itself or failure of the controller to adequately control one or more processes.

[0084] As an example, the 400 controller system may include one or more Petition 870260069218, dated 07 / 13 / 2026, p. 35 / 136 22 / 57 DTMs, such as, for example, the DTM 390 in Fig. 3. In such an example, the DTM can be an algorithm that is manually coded, etc. (e.g., a decision tree model with predefined criteria, etc.) or be a trained machine learning model (e.g., a decision tree model, one or more neural networks, etc.). As an example, a DTM can evolve from a relatively basic structure to a more complex structure that can model more "métier" as it evolves. As an example, a DTM can learn to use data so that it evolves to possess the ability to handle scenarios beyond those of a certain level of skilled operator. For example, using learning, a DTM can make inferences beyond those of a skilled operator, so that the DTM can produce parameter values ​​for ideal operating conditions that are not readily achieved (e.g., in a limited amount of time, etc.) by a skilled operator.As mentioned, a DTM can provide robust autonomous control responsive to one or more other actions taken in relation to one or more other subsystems, which may include, for example, a shutdown event where the flow at a wellhead is shut off (e.g., reduced to approximately zero).

[0085] As explained, a DTM can be data-driven and can operate responsively to the data being entered, so that the DTM can generate output. As for various types of data, the data can include fluid data acquired by a flow meter such as flow rate, flow quantity, flow characteristics, fluid characteristics, etc., and can be indicators of how a system or subsystem is behaving.

[0086] Fig. 5 shows an example of a flow meter 500 which includes circuits 510 and a sensor tube assembly U 520. As shown, the sensor tube assembly U 520 includes a pair of sensor tubes U 521 and 523 together with various conductive components 524 and 525 and various sensor components 526, 527, 528 and 529.

[0087] As mentioned, a system may include one or more types of flow meters. For example, system 300 in Fig. 3 shows several flow meters in environment 310. A flow meter may be a type of meter that can measure fluid flow and that may optionally measure one or more other types of Petition 870260069218, dated 07 / 13 / 2026, page 36 / 136 23 / 57 physical characteristics and / or phenomena (e.g., pressure, temperature, density, vibration, orientation, etc.).

[0088] As shown in Fig. 5, the flow meter 500 may include one or more sensor tubes (e.g., U-sensor tubes, etc.) 521 and 523. For example, consider the flow meter 500 as including the pair of U-sensor tubes 521 and 523 which may be provided with a split inlet flow where the U-sensor tubes 521 and 523 oscillate at a natural resonant frequency through the conductive components 524 and 525. In such an example, the flow meter 500 may include sensor components 526, 527, 528 and 529 as magnet and coil assembly pickups which may measure voltage amplitudes with respect to time (e.g., sinusoidal waves, etc.).When fluid flows in the pair of sensor tubes U 521 and 523, the Coriolis force causes each of the sensor tubes U 521 and 523 to twist in opposition to each other, resulting in a phase shift for voltage amplitudes with respect to time (e.g., phase-shifted sinusoidal waves), which are shown in graphs 532 and 534. In such an example, a time delay between phases can be measured in microseconds where the time delay is proportional to the mass flow rate (e.g., a larger time delay may correspond to a larger mass flow rate). In such an example, frequency can be measured where frequency can provide an indication of density. For example, consider Hooke's law where a tube can be a spring and the mass of a tube and fluid in it can be considered a mass coupled to the spring. In such an example, density can be estimated as being proportional to the inverse of the square of the frequency.For example, volumetric flow rate can be determined from mass flow rate and density (e.g., volumetric flow rate = mass flow rate / density).

[0089] A flow meter can be characterized by a flow rate reduction ratio (e.g., up to 100:1 or more, etc.). A flow meter can be classified according to temperature and may include one or more temperature sensors. As an example, a flow meter may be suitable for operation in a temperature range from minus 200 degrees C to plus 350 degrees C. A flow meter may include one or more types of interfaces, buses, etc. Petition 870260069218, dated 07 / 13 / 2026, page 37 / 136 24 / 57

[0090] As an example, a flow meter may include circuits that can measure flow over a range of flow rates. For example, consider a range with a lower limit that can be as low as zero and an upper limit that can be as high as, for example, 400,000 barrels per day (bpd) or more.

[0091] As an example, a flow meter can be classified with an uncertainty. For example, consider a flow uncertainty in liquids of approximately + / - 0.1 percent (e.g., + / - zero stability error). As for density, consider, for example, a density uncertainty of approximately + / - 0.0005 g / ml.

[0092] As an example, a flow meter can be constructed of several different materials, where one or more of the materials may be exposed to the fluid and considered fluid wetted. As an example, a flow meter may include one or more fluid wetted materials, such as, for example, one or more of a stainless steel and an alloy (e.g., consider 316 / 316L SST or Alloy C22).

[0093] As an example, a flow meter may be suitable for use in one or more types of hazardous areas where a hazardous area can be characterized according to one or more standards (e.g., CSA, ATEX / IECEx). In North America, hazardous locations may be defined by a combination of classes and divisions or zones, for example, as follows: Class I (a location made hazardous by the presence of flammable gases or vapors that may be present in the air in sufficient quantities to produce an explosive or flammable mixture); Class II (a location made hazardous by the presence of combustible or electrically conductive dust); Class III (a location made hazardous by the presence of easily flammable fibers or flying materials in the air, but not likely to be suspended in sufficient quantities to produce flammable mixtures); Division 1 (a location where a classified hazard exists or is likely to exist under normal conditions);Division 2 (a location where a classified hazard does not normally exist, but is possible to appear under abnormal conditions); Zone 0 (an area in which an explosive gas atmosphere is continuously present for a long period of time); Zone 1 (an area in which an explosive atmosphere is likely to occur in normal operation); Zone 2 (an area in which a; Petition 870260069218, dated 07 / 13 / 2026, page 38 / 136 25 / 57 an explosive gas atmosphere does not normally exist); etc.

[0094] As an example, a flow meter may include circuits that can perform I / O counts, for example, consider one or more independent dual pulse outputs, independent dual analog outputs, status input and status output.

[0095] As an example, a flow meter may include a serial interface, such as a Modbus RS-485 interface.

[0096] As an example, a flow meter may be compatible with one or more standards. For example, a flow meter may be compatible with the HART (Transmitter Remote Addressable Data Path) communication protocol, which is an open hybrid digital and analog industrial automation protocol. The HART approach can be used with analog instrumentation current circuits from 4 mA to 20 mA, for example, sharing a pair of wires used by an analog host system.

[0097] As an example, a flow meter may include circuits that can operate using a description language (DL). For example, consider the enhanced Electronic Device Description Language (EDDL) technology (Emerson, Saint Louis, Missouri). EDDL can provide various display technologies, for example, consider incorporating images into displays and supporting graphs and tables for calibration, configuration, etc., of control valves, variable frequency drives, etc. As an example, EDDL can be implemented to provide various storage schemes, such as, for example, persistent data storage that can allow archiving and retrieval of data from storage (e.g., measurement data, diagnostic data, calibration data, configuration data, etc.).

[0098] EDDL can provide extensions, such as those built upon the IEC 61084-2 standard, that can facilitate the design and programming of graphical display systems that can run on a variety of platforms and environments (e.g., from large Human-Machine Interfaces (HMIs) to mobile, portable devices) utilizing various common graphical display features provided by commands. Petition 870260069218, dated 07 / 13 / 2026, page 39 / 136 26 / 57 EDDL. Through extended EDDL, a common appearance can be achieved, which can facilitate more uniform integration, configuration / adjustment, operation, and diagnostics / maintenance (e.g., in an interoperable, multi-vendor environment).

[0099] As an example, a flow meter may include one or more wireless technologies, for example, for transmissions and / or receptions. For example, consider one or more infrared sensors that can provide configuration at a distance. As an example, a flow meter may include circuitry to implement an EZ-Link interface.

[00100] As an example, a flow meter may include circuits that can perform one or more self-diagnostic methods and yield results from them. For example, consider one or more diagnostic checks for pipe vibration, media pulsation, etc. As an example, a flow meter may include one or more accelerometers, one or more gyroscopes, one or more gravity meters, etc., which can measure one or more orientation, vibration, movement, etc.

[00101] As an example, a flow meter may include one or more types of end connectors. For example, consider one or more ASME flanged, DIN flanged, threaded, or tri-clover ferrules.

[00102] As an example, a DTM can be implemented using circuits (e.g., a controller unit, a controller system, etc.) that can be operatively coupled to a flow meter, for example, to at least partially receive data from the flow meter. As explained, a flow meter can include circuits that can be responsive to input. For example, consider EDDL as input to control the state of a flow meter's circuits.

[00103] As an example, a DTM may include capabilities to communicate with one or more sensors. For example, the DTM 390 in Fig. 3 may include capabilities to communicate with one or more flow meters, such as the flow meter 500 in Fig. 5. In such an example, the DTM may conduct data acquisition for input into the DTM. For example, consider the DTM requiring a time delay between the phases of the harmonic curves (e.g., sinusoidal curves) of the U-tubes, where the DTM may use the time delay specifically as a timing, which may be subject to Petition 870260069218, dated 07 / 13 / 2026, p. 40 / 136 27 / 57 one or more calculations (e.g., trend, trend rate, etc.). As an example, a DTM can access and acquire a raw time delay from a twin-pipe flow meter, where the DTM can evaluate a series of raw time delays for one or more types of behaviors, which may pertain to the fluid, fluid flow, and / or flow meter operation.

[00104] As an example, a DTM includes features for self-checking a subsystem. The DTM can communicate with one or more sensors to determine the state of those sensors. For example, consider using EDDL to request diagnostics from a flow meter. In response, a diagnostic code can be an input to the DTM, where the DTM can act to discount data, adjust data, request maintenance of the flow meter, etc. As mentioned, a DTM can be specialized to be specific to a subsystem and one or more types of operational tasks, which may include maintenance tasks, associated with the subsystem.

[00105] Fig. 6 shows an example of a system 610 that may include one or more algorithms 610 (e.g., decision tree models, etc.), one or more machine learning models 618, one or more sensors 622, one or more actuators and / or controllers 626, one or more data acquisition units 630, one or more fluid separators 634, one or more throttling manifolds 638, one or more tanks 642, one or more manifolds 646, one or more solid separators and / or manifolds 650, one or more automation systems 654 (e.g., controller systems, etc.) and one or more independent safety systems 658 (e.g., SIL-rated, etc.).

[00106] For example, system 110 may include several features of system 610. For example, system 250 may include several features of system 610. For example, system 300 may include several features of system 610.

[00107] In the example in Fig. 6, the 610 system can utilize hackcoding and / or machine learning. For example, one or more of the 610 algorithms may include one or more coded decision trees (e.g., decision tree models with coded criteria), etc., which can be developed using knowledge, experience, etc., of one or more crew members performing operations of Petition 870260069218, dated 07 / 13 / 2026, page 41 / 136 28 / 57 field in a system such as system 110 in Fig. 1, system 250 in Fig. 2, system 300 in Fig. 3, etc.

[00108] As an example, a bootstrap approach can be implemented where a specification is used for the configuration of various control units, control systems, field sensors, field actuators, etc. In such an example, the specification may be based on knowledge and experience with an automation objective. In such an approach, the specification may include a decision tree that can be operable using acquired data to make decisions regarding the values ​​of one or more parameters that can be implemented in an effort to optimize an operation. Since various types of machine learning models require sufficient training data, adherence to the specification and decision tree across multiple installations can provide sufficiently organized data generation to train one or more machine learning models.

[00109] As an example, a bootstrap approach can be layered. For instance, consider a first level as including a specification that is aligned with the equipment and operational tasks of one or more operators (e.g., one or more crew members, etc.). The implementation of the first level can provide a second level that includes generating, refining, etc., a decision tree model or other suitable decision-making algorithm that can receive input based on acquired data to output parameter values. The implementation of the second level can provide a third level that includes generating one or more trained machine learning models that are aligned with the equipment and trained to make decisions, which may include at least some decisions corresponding to the operational tasks of one or more operators.In such an example, a transfer can occur incrementally for the transition control from manual control to autonomous control.

[00110] When some amount of autonomous control is implemented, a system can provide self-tuning, which may include requesting additional learning, selecting an updated trained machine learning model, selecting a different machine learning model, etc. As an example, the Petition 870260069218, dated 07 / 13 / 2026, page 42 / 136 29 / 57 Learning can include supervised learning (e.g., using labels, etc.) and / or unsupervised learning (e.g., not using labels, etc.).

[00111] As an example, the 610 system can be configured for one or more types of operations, such as, for example, one or more surface testing operations, cleaning operations, purging operations, hydraulic fracturing operations, hydraulic fracturing plug drilling (FPDO) operations, hydraulic fracturing backflow operations, production operations, well intervention operations, production installation operations, etc.

[00112] As an example, the 610 system can be or become a self-contained surface system that is configured to self-adjust its parameters to maintain ideal process and operating conditions during the performance of one or more operations.

[00113] In the example in Fig. 6, system 610 may include a control system, such as, for example, control system 400 in Fig. 4, and may include one or more sensors including flow meters, such as, for example, flow meter 500 in Fig. 5. In such an example, system 610 may operate at least in part using data acquired by one or more sensors, in particular flow meters, to maintain ideal process and operating conditions during the performance of one or more operations involving fluid flow.

[00114] As explained, the 610 system can be applied to well testing, where a digital twin can be used. For example, the 610 system can include the DTM 390 of Fig. 3. The 610 system can provide operational control through a method that includes data acquisition and, based at least in part on at least a portion of the acquired data being fed to a trained machine learning model, generating parameter values ​​for ideal operating conditions as an output. In such an example, the parameter values ​​can be used by one or more equipment controllers to adjust setpoints in an effort to achieve ideal operating conditions.

[00115] As an example, the 610 system may include a diagnostic component and insight may include or be operationally coupled to a model of Petition 870260069218, dated 07 / 13 / 2026, page 43 / 136 30 / 57 trained machine learning (e.g., a digital twin, etc.) for the purpose of outputting useful information and / or warnings. In such an example, the diagnostic and insight component may be able to act independently and self-adjust.

[00116] As an example, a system might include one or more trained machine learning models that can receive input and generate output for autonomous adjustment of a surface system. In such an example, a trained machine learning model might be trained in a way that replicates at least in part the knowledge and operations performed by a surface crew (e.g., the jobs or métier of a surface crew). As an example, such a system might allow equipment, such as a separator, to autonomously regulate and readjust its setpoints based on actual conditions, including flow conditions. As an example, such a system might be able to regulate levels and pressures through automation and, for example, might go further by analyzing sensor data and then updating setpoints as conditions change. This regulation might be performed, at least in part, in a manner similar to that of an experienced operational crew.This approach can be implemented through training that utilizes operational tasks performed by one or more crew members in response to the acquisition of various types of sensor data. As mentioned, a DTM can be coded using the knowledge and experience of such crew members and / or trained using a machine learning model. As mentioned, a layered approach can be implemented that can transition from a coded model to a trained machine learning model for enhanced automation purposes.

[00117] As an example, the 610 system can be implemented in such a way as to allow a separator to operate autonomously and independently of one or more other pieces of equipment. In such an example, the DTM 390 as in Fig. 3 can be specialized to be a digital twin of a separator and one or more operators. In such an approach, the implementation of the DTM 390 can increase automation in the 310 environment, which can present various hazards to one or more operators. By alleviating the demand for operator presence in a hazardous environment, human risks are reduced (e.g., health, safety, and environmental risks). The DTM 390, Petition 870260069218, dated 07 / 13 / 2026, page 44 / 136 31 / 57 as mentioned, can be implemented in a way that is independent of the safety system 320 where, for example, if the DTM 390 cannot adequately handle a condition or conditions, the safety system 320 can take action (e.g., shut off the flow, shut down the equipment, etc.). As an example, the safety system 320 can be independent, although operationally coupled to the equipment implementing the DTM 390, for example, to inform the equipment that the safety system 320 has taken one or more actions. Alternatively or additionally, the equipment implementing the DTM 390 can, through data acquisition, understand or determine that one or more actions have been taken by the safety system 320 (e.g., consider a flow meter emitting a zero flow signal in response to the safety system 320 controlling the flow controller 310 to shut off the flow from a wellhead or wellheads, etc.).

[00118] As an example, DTM 390 may include safety features such that it may require one or more actions to cause a controlled shutdown of a wellhead or wellheads. In such an example, the 320 safety system may still be present, but not take action where DTM 390 is implemented in a way that allows it to take appropriate safety measures. For example, DTM 390 may be implemented in a way that allows a controlled shutdown of a well without exposing personnel to risks and without creating additional risks, such as may be triggered by an emergency shutdown.

[00119] As explained, automation can give rise to risks. As mentioned, a control system can fail due to one or more failure mechanisms. For example, consider a network failure (e.g., at the equipment level or the entire network), which could have serious consequences on a surface process if it is no longer able to adapt to changes in the process.

[00120] As an example, a network or system problem at a supervisory level (e.g., failure of acquisition or supervisory computers, network loss, etc.), which would result in unsupervised equipment unable to adjust to changes in flow or pressure, can be mitigated when a safety system is in place so that an emergency shutdown can be triggered. In such a case, Petition 870260069218, dated 07 / 13 / 2026, p. 45 / 136 32 / 57 example, however, there is the possibility of one or more undesirable equipment behaviors. For example, a shutdown in the flow controller 311 by the safety system 320 in the environment 310 may have undesirable effects on a separator. For example, the pressure accumulated in a process may maintain the flow instead of stopping immediately. In such an example, the pressure and level in the separator may get out of control, which may create spills or fire with uncontrolled burning and burner operations.

[00121] As an example, a DTM approach can maintain control over equipment during one or more procedures that may not be related to the operation of that equipment. For example, a reason for shutting down the flow at a wellhead may not be related to the conditions experienced by a separator. When a DTM is implemented for such a separator, it can maintain operational control over the separator in a way that can cause the separator equipment to operate in a manner that is not detrimental to the separator and / or that reduces the risk to other equipment and / or the environment.

[00122] As an example, a DTM approach can be implemented to have equipment handle startup, shutdown, monitoring, and continuous operations autonomously, be tolerant of supervision and loss of control, react appropriately to emergency shutdown, and act as an experienced operator, in accordance with industry best standards. Such an approach could be a digital twin with operational knowledge and experience (e.g., a digital twin with métier).

[00123] As explained, a DTM can be derived from insights and knowledge gained over the years in the industry and then implemented using one or more algorithms (e.g., models), partially or entirely at the equipment level (e.g., separator, choke manifold, etc.). In such an example, the equipment can be provided with the ability to predict changes in a process and react even in the absence of supervision and other control. In the event that a process change is of a magnitude that would exceed the capabilities of the equipment or the capabilities of the system, an emergency safety system can be used to Petition 870260069218, dated 07 / 13 / 2026, page 46 / 136 33 / 57 trigger an appropriate action where, for example, the equipment may retain the ability to self-adjust to provide a safe and controlled shutdown.

[00124] For example, in the event of a total supervisory failure, one or more components may be able to assume the supervisory function, for example, after a defined period of time, in order to continue or interrupt operations. In such an example, a DTM or DTMs may provide supervision and control of several other types of equipment in a system.

[00125] As an example, a machine learning model can be a neural network model, which can be developed based on data from a sensor data database (e.g., operational data), which may include various crew actions (e.g., valves, operation, setpoint changes, pressure, temperature, flow rate changes, etc.). For example, a machine learning model can be trained and tested on such data. As explained, where a database is continuously increasing in data volume, a model can be improved (e.g., additionally trained, retrained, etc.), which can provide the model with the ability to learn more complex patterns over time.Although various types of machine learning models and / or combinations of machine learning models are mentioned, one or more other types of machine learning models (e.g., or combinations thereof) may be appropriate for controlling the system.

[00126] Regarding the types of manually coded algorithms (e.g., models, etc.), a method may include accessing metier data, assessments and safety measures (HAZID, HAZOP, LOPA) and industry standards. As an example, from pFMEA and dFMEA, consequences, monitoring parameters, and monitoring methods can be identified for appropriate and safe actions that can be implemented, for example, in the case of a machine learning model still not being able to provide a sufficient result.

[00127] Fig. 7 shows an example of a system 700 that includes an I / O network layer 711, a control network layer 731, and a supervisory network layer 751. As shown, the I / O network layer 711 can include subsystems Petition 870260069218, dated 07 / 13 / 2026, page 47 / 136 34 / 57 712-1, 712-2, and 712-N, and a security subsystem 720, which include various components for acquiring data associated with a physical subsystem and operating the physical subsystem (e.g., control, etc.). The control network layer 731 may include a controller 730, an application server 732 (e.g., operating one or more DTMs and / or one or more other models), network equipment 734 for communications (e.g., transmissions, receptions, etc.), and a security controller 740 (e.g., operatively coupled to at least the security subsystem 720 of the I / O network layer 711, etc.). The supervisory network layer 751 may include one or more stations 750 (e.g., local station, edge station, or field station) and one or more corporate stations 754.

[00128] For example, a station can be a workstation that is a computing device or a computing system. For example, a station can be a mobile device, which can be carried by an individual, a vehicle, etc. For example, a mobile device can be transportable from location to location, system to system, subsystem to subsystem, etc., for one or more purposes, which may include, for example, local data acquisition and / or control. For example, consider issuing an instruction through a rendered graphical user interface to a mobile device display that is transmitted and / or processed through the control network layer 731 and issued through the I / O network layer 711. In such an example, a response can be issued by the equipment directly and / or indirectly to the mobile device.In simple terms, consider a local area network (e.g., wireless), a proximity-based communication protocol (e.g., Bluetooth, etc.), etc.

[00129] As an example, system 700 can provide aggregation of interactions, communications, states, conditions, etc., for one or more systems. In such an example, system 700 can be a source of information for training and / or retraining one or more machine learning models (e.g., one or more DTMs, etc.). As explained, a machine learning-based approach can improve over time, which can improve in terms of one or more aspects of prediction accuracy, ability to handle complexity, ability to scale complexity, etc. Petition 870260069218, dated 07 / 13 / 2026, page 48 / 136 35 / 57

[00130] In the 700 system, the 732 application server can serve one or more DTMs that can be, for example, deployed through the 730 controller, optionally in a way that is subsystem-based. As shown, the 740 security controller can operate with the 720 security subsystem, for example, through the 734 network equipment and / or through one or more alternative (e.g., redundant, etc.) communication channels.

[00131] In the 700 system, data can be generated at the I / O network layer level 711 at a rate of hundreds of data values ​​per second, which can be routed through network equipment. For a surface system configured for well testing, such data may include separator data, choke manifold data, downstream data, wellhead data, etc. As mentioned, a sensor may operate at a sampling rate that is on the order of milliseconds (e.g., or microseconds, etc.). As an example, the 730 controller may operate using data samples at a rate of seconds, which may be a rate intended to include data from a slower-response sensor, etc., and / or to provide a time increment that may be relevant to phenomena that may occur in a surface system.

[00132] Regarding one or more operator stations 750, they can be implemented using one or more resources (e.g., local, cloud, etc.). Regarding one or more corporate stations 754, they can be implemented using various resources that can provide access to at least the control network layer 731. As an example, a supervisory level instruction may require the construction of a DTM, the deployment of a DTM, the implementation of a DTM, etc. As an example, the application server 732 may provide the installation and / or instantiation of a DTM at a well site using a supervisory controller system (e.g., SCADA, etc.) and / or a controller unit (e.g., PLC, etc.). As mentioned, a system can deploy multiple DTMs, which can be specialized for various types of equipment in a surface system.

[00133] As mentioned, a DTM (see, for example, DTM 390, etc.) can be a model, which can be, for example, an algorithm (e.g., tree of Petition 870260069218, dated 07 / 13 / 2026, p. 49 / 136 36 / 57 hand-coded decision, etc.), a trained machine learning model (trained ML model) or a hand-coded / trained hybrid ML model that can be deployed on one or more controllers.

[00134] Regarding the types of machine learning models, consider one or more of a support vector machine (SVM) model, a nearest neighbor (KNN) model, an ensemble classifier model, a neural network (NN) model, etc. As an example, a machine learning model can be a deep learning model (e.g., deep Boltzmann machine, deep belief network, convolutional neural network, stacked autoencoder, etc.), an ensemble model (e.g., random forest, gradient boosting machine, bootstrapped aggregation, AdaBoost, stacked generalization, gradient boosted regression tree, etc.), a neural network model (e.g., radial basis function network, perceptron, backpropagation, Hopfield network, etc.).), a regularization model (e.g., ridge regression, least absolute shrinking and selection operator, elastic network, minimum angle regression), a rule system model (e.g., cubist, one rule, zero rule, repeated incremental pruning to produce error reduction), a regression model (e.g., linear regression, ordinary least squares regression, step regression, multivariate adaptive regression splines, locally estimated scatter plot smoothing, logistic regression, etc.), a decision tree model (e.g., naive Bayes, mean dependency estimators, Bayesian belief network, Gaussian naive Bayes, multinomial naive Bayes, Bayesian network), a decision tree model (e.g., classification and regression tree, iterative dichotomizer 3, C4.5, C5.0, automatic interaction detection chi-square, decision stump, conditional decision tree, M5), a dimensionality reduction model (e.g., principal component analysis, partial least squares regression, Sammon mapping, multidimensional scaling, projection search, principal component regression, partial least squares discriminant analysis, mixture discriminant analysis, quadratic discriminant analysis, regularized discriminant analysis, flexible discriminant analysis, linear discriminant analysis, etc.), an instance model (e.g., k-nearest neighbor,. Petition 870260069218, dated 07 / 13 / 2026, page 50 / 136 37 / 57 learning vector quantization, self-organizing map, locally weighted learning, etc.), a clustering model (e.g., k-means, k-medians, expectation maximization, hierarchical clustering, etc.), etc.

[00135] Above, an autoencoder is mentioned, which may include an encoder that is trained in combination with a decoder. In such an example, unsupervised training can be used where the encoder and decoder are trained so that the output properly matches the input. As an example, as mentioned, information in a hidden layer can be used for one or more purposes. An encoder can generate such information, which can be referred to as information or embeddings in a latent space. In such an example, the information or embedding can be a compressed representation of input, which can characterize the input. For example, a large array or vector of information (e.g., data, etc.) can be compressed using a trained encoder to generate a smaller array or vector, which may be suitable for use in one or more control schemes, comparisons, etc.

[00136] An autoencoder (or auto-encoder) can be characterized as a type of artificial neural network that can learn efficient data encodings in an unsupervised manner. For example, an autoencoder can process data to learn a representation (encoding) for the data (e.g., for dimensionality reduction, etc.). Such an approach may involve training the encoder to ignore types of noise (e.g., signal noise, etc.). As explained, a decoder can be trained in combination with an encoder. One or more of several types of autoencoders can be used (e.g., sparse, noiseless, contractive, variational, etc.), which can provide effective learning of representations for classification tasks, generative models, etc. As an example, an autoencoder can be used for recognition (e.g., data-based, image-based, etc.), semantic meaning of symbols, etc.

[00137] As an example, a DTM may include one or more characteristics of an autoencoder. For example, consider an autoencoder trained using sensor data, control data, etc., which may include metadata. Petition 870260069218, dated 07 / 13 / 2026, page 51 / 136 38 / 57 associated. In such an example, a DTM can provide classification of one or more types of conditions, control scenarios, etc., based on the input, so that control can be effected, as desired, efficiently, optionally automatically. In particular, such a DTM can include sensor data as input and output control data (e.g., including one or more setpoints) that can be implemented in the subsystems through appropriate control equipment (e.g., one or more controllers, one or more actuators, etc.).

[00138] As an example, a machine model, which can be a machine learning model, can be built using a computational framework with a library, a toolbox, etc., such as, for example, those of the MATLAB framework (MathWorks, Inc., Natic, Massachusetts). The MATLAB framework includes a toolbox that provides supervised and unsupervised machine learning algorithms, including support vector machines (SVMs), boosted and bagged decision trees, k-nearest neighbor (KNN), k-means, K-medoids, hierarchical clustering, Gaussian mixture models, and hidden Markov models. Another toolbox of the MATLAB framework is the Deep Learning Toolbox (DLT), which provides a framework for designing and implementing deep neural networks with algorithms, pre-trained models, and applications.DLT provides convolutional neural networks (ConvNets, CNNs) and long short-term memory (LSTM) networks to perform classification and regression on images, time series, and text data. DLT includes features for building network architectures such as generative adversarial networks (GANs) and Siamese networks using custom training circuits, shared weights, and automatic differentiation. DLT provides several other frameworks for model switching.

[00139] Fig. 8 shows an example of an 800 model, which is a decision tree model. As shown, the 800 model includes 810 input and 830 output. As an example, the 800 model can be a generative model of induction rules as derived from acquired data and operational performance by one or more operators. An ideal decision tree can be defined as a tree that represents the majority of the data while minimizing the number of levels (e.g., questions). As an example, one or more Petition 870260069218, dated 07 / 13 / 2026, page 52 / 136 39 / 57 techniques can be implemented to generate an ideal tree (e.g., ID3 / 4 / 5, CLS, ASSISTANT, CART, etc.).

[00140] Fig. 9 shows an example of a 900 machine learning model that can be a neural network (NN). As shown, the 900 model can include an input layer, one or more hidden layers, and an output layer. For example, input can be received through the input layer to generate information in the hidden layer and to generate information in the output layer. For example, information in at least one of a hidden layer and one output layer can be used for one or more purposes. As mentioned, an autoencoder can provide the generation of representations (embeddings) in a latent space where “latent” can refer to “hidden”.

[00141] A neural network (NN) can include neurons and connections where each connection provides the output of one neuron as an input to another neuron. Each connection can be assigned a weight representing its relative importance. A given neuron can have multiple input and output connections. An NN can include a propagation function that calculates the input to a neuron from the outputs of its predecessor neurons and its connections as a weighted sum. As an example, a bias term can be added to the propagation result.

[00142] As an example, neurons can be organized into multiple layers, particularly in deep learning NNs. As explained, the layer that receives external data can be an input layer, and the layer that produces a result or results can be an output layer. As an example, an NN can be fully connected where each neuron in one layer connects to every neuron in the next layer. As an example, an NN can utilize clustering, where a group of neurons in one layer connects to a single neuron in the next layer, thus reducing the number of neurons in that layer. As an example, an NN can include connections that form a directed acyclic graph (DAG), which can define feedforward type networks. Alternatively, an NN can allow connections between neurons in the same layer or in previous layers (e.g., a recurrent network). Petition 870260069218, dated 07 / 13 / 2026, page 53 / 136 40 / 57

[00143] As an example, an NN can be a recurrent neural network (RNN), which is a class of artificial neural networks (ANNs) where the connections between nodes can form a directed graph (DG) over a temporal sequence. An RNN can exhibit temporal dynamic behavior. Compared to feedforward neural networks, an RNN can use its internal state (memory) to process variable-length sequences of inputs.

[00144] As explained, one or more neural network models can be developed based on data (e.g., datasets, etc.) from one or more databases, live streams, etc., to sensor data (e.g., operational data) and associated types of crew actions (e.g., actions regarding one or more valves, operations, setpoint changes, pressures, temperatures, flow rate changes, etc.) where, for example, individuals performing such crew actions may include individuals who have been trained and tested (e.g., for system and / or one or more subsystem operations). In such an example, the data can grow continuously where one or more such models can be improved (e.g., trained, retrained, etc.) so that, for example, more complex patterns can be learned over time.In such an example, more complex patterns may include patterns regarding physical phenomena that may involve interactions between one or more subsystems (e.g., complex interdependencies, etc.), for example, regarding how such one or more subsystems can be controlled.

[00145] As an example, an RNN can be characterized as a finite impulse model or as an infinite impulse model, either of which can exhibit temporal dynamic behavior. A finite impulse recurrent network can be a directed acyclic graph (DAG) that can be unrolled and replaced by a strictly feedforward neural network, while an infinite impulse recurrent network can be a directed cyclic graph (DAG) that cannot be unrolled.

[00146] An RNN, whether as a finite pulse or an infinite pulse recurrent network, may include additional stored states where the storage may be under direct control by the RNN. As an example, the storage may also be replaced by another network or graph (e.g., consider time delays, Petition 870260069218, dated 07 / 13 / 2026, page 54 / 136 41 / 57 feedback circuits, etc.). Such controlled states may be referred to as closed states or closed memories and may be part of a long-term memory (LSTM) approach, a closed recurrent units (GRUs) approach, etc. (e.g., consider a feedback-type neural network).

[00147] An LSTM can be part of a deep learning system that may, for example, aim to solve the vanishing gradient problem. LSTM can be augmented by recurrent gates (e.g., forgetting gates). LSTM can reduce the risks of backpropagating vanishing or explosion errors. For example, errors can flow backward through multiple virtual layers unfolded in space. LSTM can learn tasks that require memory of one or more events that occurred a number of discrete time steps before. An LSTM approach can be employed with various time types, even with long delays between specific events. An LSTM approach can be employed where signals may include a range of frequencies (e.g., mixture of low and high frequency components, etc.).

[00148] As an example, a machine model can utilize stacks of LSTM RNNs, which can be, for example, trained via Connectionist Temporal Classification (CTC) to find an RNN weight matrix that maximizes the probability of the labeling sequences in a training set, given the matching input sequences. CTC can achieve alignment and recognition.

[00149] As mentioned, a machine model can include one or more closed recurrent units (GRUs). A GRU can be a locking mechanism in an RNN. As an example, a GRU can be used in addition to or as an alternative to an LSTM; noting that a GRU approach may have fewer parameters than an LSTM approach, as a GRU may be without an output gate.

[00150] As an example, a machine learning model can be trained to handle data that can be synchronous and / or data that can be asynchronous. For example, a subsystem can operate in a way that can generate data (e.g., sensor data, etc.) that are available according to a Petition 870260069218, dated 07 / 13 / 2026, page 55 / 136 42 / 57 Synchronous data transmission technique where, for example, the data is accompanied by timing signals (e.g., generated by an electronic clock) to ensure that the transmitter and receiver are in step (synchronized) with each other. In such an example, the data may be transmitted in blocks (e.g., frames or packets) spaced by fixed time intervals.

[00151] As for asynchronous data, it can be transmitted in spurts where a transmission technique might include inserting a start bit before each data character and a stop bit at its end to inform a receiver where the data “spurt” begins and ends.

[00152] Although the previous examples describe various aspects of transmission, synchronous can refer to data that is continuously transmitted, and asynchronous can refer to data that is not continuously transmitted. For example, when an alarm condition is reached, this alarm condition may cause the transmission of data, which may differ from the continuously transmitted data. As another example, an alarm condition may trigger the transmission of data that was not previously transmitted. In such an approach, an alarm condition may initiate a synchronous data stream. As mentioned, in synchronous transmission, a transmission time interval may be constant; whereas, in asynchronous transmission, a transmission time interval may be different from constant (e.g., different intervals, random intervals, etc.).In a synchronous transmission mode, data can be transmitted without a "gap" between the data; whereas in an asynchronous transmission mode, the data can be separated by a "gap," which, as mentioned, can be irregular.

[00153] As an example, a system and / or a subsystem may generate data that may be available regularly or irregularly. As an example, a machine learning model may include an architecture that can handle regular and / or irregular data. As an example, consider a machine learning model that may receive inputs and may generate one or more outputs where one output may be a prediction based at least in part on the received inputs.

[00154] As an example, a machine learning model can be or Petition 870260069218, dated 07 / 13 / 2026, p. 56 / 136 43 / 57 include a multivariate time series regression model. For example, in multiple regression, a response variable Y may be related to a number of predictor variables, which can be projected by a number k. For example, consider the following equation where t is time: Y = b0 + b1X1,t + . . . + bkXk,t + noise

[00155] In a multivariate time series regression, there may be m response time series variables related to the predictor time series variables k.

[00156] As an example, a machine learning model may include one or more features of a significance-compensating convolutional neural network (SOCNN) that has a deep convolutional network architecture for multivariate time series regression.

[00157] For example, consider a model rooted in an autoregressive (AR) model that includes gating mechanisms used in recurrent neural networks (RNNs). In such an example, an AR-like weighting system can be implemented where a final predictor is obtained as a weighted sum of sub-predictors, while the weights are data-dependent functions learned through a convolutional network. As an example, such an architecture might be suitable for scenarios that may include asynchronous time series. As an example, a machine learning model might operate to predict ahead in a manner conditioned on a number of past observations.

[00158] As an example, a machine learning model might include training using data from multiple sources. For example, individual operators might be individual sources of information, which could provide such information at irregular times. In such an example, operator A might provide information (e.g., analysis, control commands, etc.) less frequently than operator B, although the underlying dynamics of a process might include similarities. In such an example, one operator might experience fewer problems than another, where a problem might be due to an incorrect prediction, a late prediction, etc. As an example, data from multiple sources could refer to how an operator Petition 870260069218, dated 07 / 13 / 2026, page 57 / 136 44 / 57 responds to a shutdown. For example, consider an ESD system that causes a shutdown of the flow in a well where the operator takes one or more actions to deal with the risk of one or more damaging events in a subsystem (e.g., a separator, etc.). In such an example, a response variable might be a control variable related to the subsystem. As an example, a trained machine learning model might receive input and output of a response variable that is a control variable related to the subsystem where such a trained machine learning model might produce the response variable as a prediction (e.g., before a demand to implement appropriate control). Such an approach might involve a digital twin that is a “twin” of one or more things (e.g., equipment, a human operator, etc.).

[00159] As mentioned, an operator may be responsible for one or more pieces of equipment, subsystems, systems, etc. As mentioned, a separator may be a subsystem where data is available for how different operators operate such a subsystem responsive to the generated data, which may include synchronous and / or asynchronous data. For example, the data may be at a particular frequency, random, etc. For example, the data may be generated, transmitted, received, available, etc., on one or more time scales, one or more frequencies, one or more resolutions, one or more signal-to-noise ratios, etc.

[00160] In a paper by Binkowski et al., entitled Autoregressive Convolutional Neural Networks for Asynchronous Time Series, ICML 2017 Time Series Workshop, Sydney, Australia, 2017, which is incorporated by reference in this document, different sources of financial data were considered where the value of an asset is observed differently by each of the different sources, which may observe irregularly over time, different biases, different noise, etc. However, the sources may be somewhat correlated and may include leading-lagging relationships. And, it is possible that the sources may perform adequately so that their behavior is worth making an effort to “capture” by coupling linear models with deep neural networks that can learn non-linear relationships. Petition 870260069218, dated 07 / 13 / 2026, page 58 / 136 45 / 57

[00161] As mentioned, a machine learning model can be trained using training data from or derived from human operators, where such data may optionally include data from or derived from machine operators (e.g., automated controller, etc.). As an example, a method might include building a first model, operating equipment according to the first model, acquiring data, and using the data to train a second model. In such an example, there might be a progression from human control to machine control. As for the first model, it can be built using data from multiple sources, which may include multiple operators. As operations progress, one or more other models (e.g., or the same model) can be trained using data from these progressed operations.In such an approach, human operator data can help initiate the development of an automated controller that is based, at least in part, on a machine learning model.

[00162] Regarding a machine learning model (ML model), as explained, such a model can be a neural network model (NN model). As an example, a trained ML model can be used to control one or more subsystems. Various types of data can be acquired and optionally stored, which can provide training for one or more ML models and / or for offline analysis, etc. For example, air control parameters emitted by a trained NN model can be stored in digital storage for later analysis, which may include additional training, training of a different ML model, etc.

[00163] As an example, the TENSORFLOW framework (Google LLC, Mountain View, CA) can be implemented, which is an open-source software library for data flow programming that includes a symbolic math library, which can be implemented for machine learning applications that may include neural networks. As an example, the CAFFE framework can be implemented, which is a DL framework developed by Berkeley AI Research (BAIR) (University of California, Berkeley, California). As another example, consider the SCIKIT platform (e.g., scikit-learn), which uses the PYTHON programming language. As an example, a framework such as the APOLLO AI framework can be used (APOLLO.AI). Petition 870260069218, dated 07 / 13 / 2026, page 59 / 136 46 / 57 GmbH, Germany). As an example, a framework like the PYTORCH framework can be used (Facebook AI Research Lab (FAIR), Facebook, Inc., Menlo Park, California).

[00164] As an example, a training method can include several actions that can operate on a dataset to train an ML model. For example, a dataset can be divided into training data and test data, where the test data can provide evaluation. A method can include parameter cross-validation and best parameters, which can be provided for model training.

[00165] The TENSORFLOW framework can run on a variety of CPUs and GPUs (with optional CUDA (NVIDIA Corp., Santa Clara, California) and SYCL extensions (The Khronos Group Inc., Beaverton, Oregon) for general-purpose computing on graphics processing units (GPUs)). TENSORFLOW is available on 64-bit LINUX, MACOS (Apple Inc., Cupertino, California), WINDOWS (Microsoft Corp., Redmond, Washington), and mobile computing platforms, including Android (Google LLC, Mountain View, California) and iOS (Apple Inc.) based operating system platforms.

[00166] TENSORFLOW calculations can be expressed as state data flow graphs; noting that the name TENSORFLOW derives from the operations that such neural networks perform on multidimensional data matrices. Such matrices can be referred to as “tensors”.

[00167] Fig. 10 shows a 1000 architecture of a framework, such as the TENSORFLOW framework. As shown, the 1000 architecture includes several features. For example, in the terminology of the 1000 architecture, a client can define a computation as a data flow graph and, for example, can start the execution of the graph using a session. For example, a distributed master can prune a specific subgraph of the graph, as defined by the arguments to “Session.run()”; partition the subgraph into several pieces that are executed in different processes and devices; distribute the graph pieces to worker services; and start the execution of the graph piece by the worker services. As for the worker services (for example, one per task), for example, they can Petition 870260069218, dated 07 / 13 / 2026, page 60 / 136 47 / 57 Schedule the execution of graph operations using kernel implementations appropriate to the available hardware (CPUs, GPUs, etc.) and, for example, send and receive operation results to and from other worker services. As for the kernel implementations, these can, for example, perform calculations for individual graph operations.

[00168] Fig. 11 shows an example of a workflow 1100 that includes a provision block 1110 to provide a specification regarding the equipment and operations of a system, a generation block 1120 to generate a model using system data adhering to the specification (e.g., a decision tree model, etc.), a generation block 1130 to generate a machine learning model trained using data acquired from the model implementation to control the system; and a control block 1140 to control the system using the trained machine learning model.

[00169] In the example in Fig. 11, the descriptive report may include subsystem information regarding specific subsystems of the system, whereby the subsystems are controllable through parameters in which an operator skilled in the art, in a manual operating mode, would set various parameter values ​​in an effort to optimize the operation of the subsystem. As an example, in Fig. 11, the model (e.g., an algorithm, etc.) could be a decision tree model. As an example, where the data is sufficient for the generation of a machine learning model, workflow 1100 could proceed to block 1130, optionally without executing block 1120 (e.g., without first controlling the system according to the algorithm, etc.). As an example, the machine learning model could be a feedforward neural network that receives data as input and produces one or more parameter values ​​in a probabilistic manner.In such an example, one or more parameter values ​​can be communicated to one or more components of a system for the purpose of controlling at least a portion of the system.

[00170] Fig. 12 shows an example of a 1200 method and an example of a 1290 system. As shown, the 1200 method includes a 1210 acquisition block to acquire data during the operation of a surface system that handles fluid. Petition 870260069218, dated 07 / 13 / 2026, page 61 / 136 48 / 57 using subsystems; a generation block 1220 to generate a parameter value for one of the subsystems using at least a portion of the acquired data and a model that models the subsystem using historical data that includes operational tuning data for condition-responsive operational adjustments; and a control block 1230 to automatically control one of the subsystems using the parameter value. In such an example, the model might be a machine learning model that has been trained using historical data (e.g., including historical acquired sensor data and historical operational tuning data) and that allows output of the parameter value based on the acquired data input.

[00171] Method 1200 is shown as including several blocks of computer-readable storage media (CRM) 1211, 1221, and 1231 that may include processor-executable instructions that may instruct a computing system, which may be a control system, to perform one or more of the actions described in relation to method 1200.

[00172] In the example in Fig. 12, the system 1290 includes one or more information storage devices 1291, one or more computers 1292, one or more networks 1295, and instructions 1296. As for one or more computers 1292, each computer may include one or more processors (e.g., or processing cores) 1293 and memory 1294 to store instructions 1296, for example, executable by at least one of the one or more processors 1293 (see, for example, blocks 1211, 1221, 1231, and 1241). As an example, a computer may include one or more network interfaces (e.g., wired or wireless), one or more graphics cards, a display interface (e.g., wired or wireless), etc.

[00173] As an example, method 1200 can be a workflow that can be implemented using one or more frameworks that may be within a framework environment. As an example, system 1290 can include local and / or remote resources. For example, consider a browser application running on a client device as a local resource relative to a user of the browser application, and a cloud-based computing device as a remote resource relative to the user. In such an example, the user can interact with the device. Petition 870260069218, dated 07 / 13 / 2026, page 62 / 136 49 / 57 client via the browser application where information is transmitted to the cloud-based computing device(s) and where information can be received in response and rendered to a display operationally coupled to the client device (e.g., via services, APIs, etc.).

[00174] As an example, a method might include acquiring data during the operation of a surface system that handles fluid using subsystems; generating a parameter value for one of the subsystems using at least a portion of the data and a model that models the subsystem using historical data that includes operational tuning data for condition-responsive operational adjustments; and automatically controlling one of the subsystems using the parameter value.

[00175] For example, a model may be or include a decision tree model. For example, a model may be or include a machine learning model. For example, a model may be or include a neural network model. For example, a model may be or include a métier digital twin model. For example, a model may be a hybrid model that includes a coded model and / or portion and a trained model and / or portion.

[00176] As an example, a method might include training a model using historical data and associated historical parameter values ​​to learn weights as data-dependent functions, where a generated parameter value is based at least in part on the output of the trained model, where the output depends on the learned weights being applied to variables of the trained model (e.g., consider a regression model type, etc.).

[00177] As an example, operational adjustment data may include manual adjustment data for a modeled subsystem. For example, consider decisions made by one or more operators responding to various types of input (e.g., sensor data, standard operating procedures, etc.).

[00178] As an example, a method might include automatically controlling a subsystem using a controller located in an area classified as a hazardous area.

[00179] As an example, a model can be used to generate a value of Petition 870260069218, dated 07 / 13 / 2026, page 63 / 136 50 / 57 is a parameter responsive to acquired data, indicating a shutdown of fluid flow to a system. For example, consider automatically controlling a subsystem using the parameter value to operate the subsystem and reduce the risk of damage to the subsystem.

[00180] As an example, a subsystem may be a subsystem of a well testing system, for handling fluid extracted from at least one wellbore, where the subsystem may include one or more of a collector, a choke collector, a burner and a separator.

[00181] As an example, a subsystem might include a separator to separate fluid extracted from at least one wellbore during a well test where, for example, the separator includes at least one flow meter. In such an example, a parameter value might correspond to a state of a flow valve in the subsystem.

[00182] As an example, the acquired data may be indicative of one or more pressures, temperatures, and flow rates in a subsystem or, for example, multiple subsystems where some interdependence exists.

[00183] As an example, the acquired data may be indicative of a pressure being greater than a desired pressure. In such an example, a safety assessment may be performed, for example, depending on how much greater the pressure may be than a desired pressure. As mentioned, one or more types of control schemes may include setpoint control. For example, a control action may be for a change in a setpoint and / or for a change that aims to help maintain a setpoint; noting that several other types of control actions may be performed using one or more outputs from a model or models.

[00184] As an example, a parameter value may correspond to a setpoint value for fluid flow in the subsystem. As an example, a parameter value may correspond to a setpoint value for pressure in the subsystem.

[00185] As an example, a method may consider conditions, which may include one or more of the ideal operating conditions, emergency operating conditions, and diagnostic conditions for diagnostics of at least one flow meter. Petition 870260069218, dated 07 / 13 / 2026, page 64 / 136 51 / 57

[00186] As an example, the acquired data may be indicative of a pressure being greater than a desired pressure. As an example, a parameter value may correspond to a setpoint value for fluid flow in a subsystem. As an example, a parameter value may correspond to a setpoint value for pressure in a subsystem.

[00187] As an example, a method may include the use of at least a portion of acquired data and a model that models a subsystem using historical data, which includes operational tuning data for condition-responsive operational adjustments, where conditions may include ideal operating conditions, emergency operating conditions, and / or diagnostic conditions for diagnosing at least one piece of equipment in a subsystem (e.g., consider a flow meter, etc.).

[00188] As an example, a method might include acquiring data during the operation of a surface system that handles fluid using subsystems; generating a parameter value for one of the subsystems using at least a portion of the data and a model that models the subsystem using historical data that includes operational tuning data for condition-responsive operational adjustments; automatically controlling one of the subsystems using the parameter value; acquiring additional data during the operation of the surface system and updating the model using at least a portion of the additional data, where the update includes at least one relearning of at least one existing weight and learning at least one new weight, where the at least one new weight increases the complexity of the model. For example, a method might include continuous learning, which could be partly through feedback regarding machine action and / or partly through feedback regarding human action.For example, consider a method that can be implemented with some amount of human supervision, which can be decreased over time as a model or models learn at least how the human action can be performed. As explained, a model or models can learn complexity that may be beyond what is expected of a human based on the human's training. For example, a model or models can learn complexities regarding interdependencies between... Petition 870260069218, dated 07 / 13 / 2026, page 65 / 136 52 / 57 subsystems, which can improve the overall control (e.g., in terms of predictability and / or efficiency) of a system, such as, for example, system 110 in Fig. 1, system 250 in Fig. 2, system 300 in Fig. 3, etc.

[00189] As an example, a system may include a processor; processor-accessible memory; processor-executable instructions stored in memory and executable by the processor to instruct the system to: acquire data during the operation of a surface system that handles fluid using subsystems; generate a parameter value for one of the subsystems using at least a portion of the data and a model that models the subsystem using historical data that includes operational tuning data for condition-responsive operational adjustments; and automatically control the subsystem using the parameter value.

[00190] As an example, one or more computer-readable storage media may include executable computer instructions to instruct a computing system to: acquire data during the operation of a surface system that handles fluid using subsystems; generate a parameter value for one of the subsystems using at least a portion of the data and a model that models the subsystem using historical data that includes operational tuning data for condition-responsive operational adjustments; and automatically control the subsystem using the parameter value.

[00191] As an example, a computer program product may include computer executable instructions to instruct a computer system to execute one or more methods, such as, for example, method 1100 of Fig. 11, method 1200 of Fig. 12, etc.

[00192] As an example, a method can be implemented in part using machine-readable media (CRM), for example, as a module, a block, etc., which includes information such as instructions suitable for execution by one or more processors (or processor cores) to instruct a device or computing system to perform one or more actions. As an example, a single medium can be configured with instructions to allow, at least in part, the performance of several actions of a method. As an example, a machine-readable medium (CRM) can be Petition 870260069218, dated 07 / 13 / 2026, page 66 / 136 53 / 57 a computer-readable storage medium (e.g., a non-transient medium) that is not a carrier wave.

[00193] According to one embodiment, one or more computer-readable media may include computer-executable instructions to instruct a computer system to send information to control a process. For example, such instructions may provide output for a detection process, an injection process, a drilling process, an extraction process, an extrusion process, a pumping process, a heating process, a burning process, an analysis process, etc.

[00194] In some embodiments, a method or methods may be executed by a computing system. Fig. 13 shows an example of a 1300 system that may include one or more 1301-1, 1301-2, 1301-3, and 1301-4 computing systems, which may be operatively coupled by means of one or more 1309 networks, which may include wired and / or wireless networks.

[00195] As an example, a system may include an individual computer system or an arrangement of distributed computer systems. In the example of Fig. 13, the computer system 1301-1 may include one or more modules 1302, which may be or include processor-executable instructions, for example, executables to perform various tasks (e.g., receive information, request information, process information, simulation, output information, etc.).

[00196] As an example, a module may run independently, or in coordination with, one or more processors 1304, which is (or are) operatively coupled to one or more storage media 1306 (e.g., wired, wireless, etc.). As an example, one or more of the one or more processors 1304 may be operatively coupled to at least one of one or more network interfaces 1307. In such an example, the computer system 1301-1 may transmit and / or receive information, for example, through one or more networks 1309 (e.g., consider one or more of the Internet, a private network, a cellular network, a satellite network, etc.).

[00197] For example, computer system 1301-1 can receive and / or Petition 870260069218, dated 07 / 13 / 2026, page 67 / 136 54 / 57 transmit information to one or more other devices, which may be or include, for example, one or more of the computer systems 1301-2, etc. A device may be located in a physical location that differs from that of the computer system 1301-1. For example, a location may be, for example, a processing facility location, a data center location (e.g., server farm, etc.), a probe location, a well location, a wellhead location, etc.

[00198] As an example, a processor may be or include a microprocessor, microcontroller, processor module or subsystem, programmable integrated circuit, programmable gate array, or other control or computing device.

[00199] As an example, storage media 1306 can be implemented as one or more computer-readable or machine-readable storage media. As an example, storage can be distributed within and / or between multiple internal and / or external compartments of a computing system and / or additional computing systems.

[00200] As an example, a storage medium or storage media may include one or more different forms of memory, including semiconductor memory devices, such as dynamic or static random access memories (DRAMs or SRAMs), erasable and programmable read-only memories (EPROMs), electrically erasable and programmable read-only memories (EEPROMs) and flash memories, magnetic disks, such as hard disks, floppy disks and removable disks, other magnetic media including tapes, optical media, such as compact discs (CDs) or digital video discs (DVDs), Blu-ray discs or other types of optical storage, or other types of storage devices.

[00201] As an example, a storage medium or media may be located on a machine executing machine-readable instructions, or located at a remote location from which machine-readable instructions may be downloaded onto a network for execution.

[00202] For example, various components of a system, such as a computer system, can be implemented in hardware, software, or a Petition 870260069218, dated 07 / 13 / 2026, page 68 / 136 55 / 57 combination of hardware and software (e.g., including firmware), including one or more signal processing and / or application-specific integrated circuits.

[00203] As an example, a system may include a processing appliance that may be or include general-purpose processors or application-specific chips (e.g., or chipsets), such as ASICs, FPGAs, PLDs, or other appropriate devices.

[00204] Fig. 14 shows components of a computing system 1400 and a networked system 1410 that includes a network 1420. The system 1400 includes one or more processors 1402, memory and / or storage components 1404, one or more input and / or output devices 1406, and a bus 1408. According to one embodiment, instructions may be stored in one or more computer-readable media (e.g., memory / storage components 1404). Such instructions may be read by one or more processors (e.g., the processor(s) 1402) via a communication bus (e.g., the bus 1408), which may be wired or wireless. One or more processors may execute such instructions to implement (fully or partially) one or more attributes (e.g., as part of a method). A user can view the output and interact with a process through an I / O device (for example, device 1406).According to one embodiment, a computer-readable medium can be a storage component, such as a physical memory storage device, for example, a chip, a chip in a package, a memory card, etc.

[00205] According to one embodiment, the components may be distributed, as in the 1410 network system. The 1410 network system includes 1422-1, 1422-2, 1422-3, ... 1422-N components. For example, the 1422-1 components may include the 1402 processor(s) while the 1422-3 component(s) may include memory accessible by the 1402 processor(s). In addition, the 1422-2 component(s) may include an I / O device for display and, optionally, interaction with a method. The network may be or include the Internet, an intranet, a cellular network, a satellite network, etc.

[00206] For example, a device can be a mobile device that Petition 870260069218, dated 07 / 13 / 2026, page 69 / 136 56 / 57 includes one or more network interfaces for information communication. For example, a mobile device may include a wireless network interface (e.g., operable via IEEE 802.11, ETSI GSM, BLUETOOTH, satellite, etc.). As an example, a mobile device may include components such as a main processor, memory, a screen, screen graphics circuitry (e.g., optionally including touch and gesture circuitry), a SIM slot, audio / video circuitry, motion processing circuitry (e.g., accelerometer, gyroscope), wireless LAN circuitry, smart card circuitry, transmitter circuitry, GPS circuitry, and a battery. For example, a mobile device may be configured as a cell phone, a tablet, etc. As an example, a method may be implemented (e.g., wholly or in part) using a mobile device. As an example, a system may include one or more mobile devices.

[00207] For example, a system can be a distributed environment, for example, a so-called “cloud” environment where various devices, components, etc. interact for purposes of data storage, communications, computing, etc. As an example, a device or system may include one or more components for communicating information over one or more Internet networks (e.g., where communication occurs via one or more Internet protocols), a cellular network, a satellite network, etc. As an example, a method may be implemented in a distributed environment (e.g., wholly or partly as a cloud-based service).

[00208] For example, information can be input from a display (e.g., consider a touchscreen), sent to a display, or both. For example, information can be sent to a projector, a laser device, a printer, etc., so that the information can be visualized. For example, information can be produced stereographically or holographically. Regarding a printer, consider a 2D or 3D printer. For example, a 3D printer can incorporate one or more substances that can be produced to construct a 3D object. For example, data can be fed to a 3D printer to construct a 3D representation of an underground formation. For example, layers can be constructed in 3D (e.g., horizons, etc.), geobodies Petition 870260069218, dated 07 / 13 / 2026, pp. 70 / 136 57 / 57 constructed in 3D, etc. For example, holes, fractures, etc., can be constructed in 3D (e.g., as positive structures, as negative structures, etc.).

[00209] Although only a few examples have been described in detail above, those skilled in the art will readily understand that many modifications to the examples are possible. Consequently, all such modifications are intended to be included within the scope of this disclosure, as defined in the following claims. In the claims, clauses of the type "means plus function" are intended to cover the structures described in this document as performing the cited function and not only structural equivalents, but also equivalent structures. Thus, although a nail and a screw may not be structural equivalents in that a nail employs a cylindrical surface to fasten pieces of wood together, whereas a screw employs a helical surface, in the context of fastening pieces of wood together, a nail and a screw may be equivalent structures. Petition 870260069218, dated 07 / 13 / 2026, page 71 / 136

Claims

1 / 3 CLAIMS 1. Method for controlling surface systems for hydrocarbon reservoirs, characterized in that it comprises: acquiring data during the operation of a surface system (110, 250, 300, 610, 700) that handles fluid using subsystems (120, 122, 124, 126, 712-1, 712-2, 712N); generate a parameter value for one of the subsystems (120, 122, 124, 126, 712-1, 712-2, 712-N) using at least a portion of the acquired data and a model (800, 900) that models the subsystem using historical data comprising operational adjustment data for condition-responsive operational adjustments, wherein the model generates the parameter value responsive to the acquired data being indicative of a fluid flow shutdown to the system; and automatically control one of the subsystems (120, 122, 124, 126, 712-1, 7122, 712-N) using the parameter value.

2. Method according to claim 1, characterized in that the model (800, 900) comprises a decision tree model (800), a machine learning model (618, 900) trained using historical data, or a neural network model (900) trained using historical data.

3. Method, according to claim 1, characterized in that it comprises training the (800, 900) model using historical data and associated historical parameter values ​​to learn weights as data-dependent functions, wherein the generated parameter value is based at least in part on the output of the trained model, wherein the output depends on the learned weights being applied to variables of the trained model.

4. Method according to claim 1, characterized in that the operational adjustment data comprises manual adjustment data for a modeled subsystem (120, 122, 124, 126, 712-1, 712-2, 712-N).

5. Method according to claim 1, characterized in that the subsystem (120, 122, 124, 126, 712-1, 712-2, 712-N) is located in an area classified as a hazardous area (301), wherein the automatic control of the subsystem Petition 870260069218, dated 13 / 07 / 2026, page 72 / 136 2 / 3 uses a controller (330, 400, 730) located in the hazardous area (301).

6. Method according to claim 1, characterized in that the subsystem (120, 122, 124, 126, 712-1, 712-2, 712-N) is a subsystem of a well testing system (250), for handling fluid (252) extracted from at least one wellbore, wherein the subsystem comprises one or more of a collector (130, 134, 146-1, 146-2, 266, 274, 280, 290, 646), a choke collector (262, 638), a burner (152, 2761, 276-2) and a separator (142, 270, 634).

7. Method according to claim 1, characterized in that the subsystem comprises a separator (142, 270, 634) for separating fluid (252) extracted from at least one wellbore during a well test, wherein the separator (142, 270, 634) comprises at least one flow meter (138, 264, 500), and wherein the parameter value corresponds to a state of a flow valve of the subsystem (120, 122, 124, 126, 712-1, 712-2, 712-N).

8. Method according to claim 1, characterized in that the acquired data are indicative of one or more pressures, temperatures and flow rates in the subsystem (120, 122, 124, 126, 712-1, 712-2, 712-N).

9. Method according to claim 1, characterized in that the acquired data are indicative of a pressure being greater than a desired pressure.

10. Method according to claim 1, characterized in that the parameter value corresponds to a setpoint value for fluid flow in the subsystem (120, 122, 124, 126, 712-1, 712-2, 712-N), or to a setpoint value for pressure in the subsystem.

11. Method according to claim 1, characterized in that the conditions comprise one or more of the ideal operating conditions, emergency operating conditions and diagnostic conditions for diagnostics of at least one flow meter (138, 264, 500).

12. Method according to claim 1, characterized in that it comprises acquiring additional data during the operation of the surface system (110, Petition 870260069218, dated 07 / 13 / 2026, p. 73 / 136 3 / 3 250, 300, 610, 700) and updating the model (800, 900) using at least a portion of the additional data, wherein the update comprises at least one of relearning at least one existing weight and learning at least one new weight, wherein the at least one new weight increases the complexity of the model.

13. System (110, 250, 300, 610, 700, 1290, 1300, 1400), characterized in that it comprises: a processor (112, 1293, 1304, 1402); memory (114, 1294, 1306, 1404) accessible to the processor; processor-executable instructions (116, 1296, 1302) stored in memory and executable by the processor to instruct the system to perform the steps of the method as defined in any one of claims 1 to 12.

14. Computer-readable storage medium, characterized in that it comprises executable computer instructions to instruct a computing system to perform the steps of the method as defined in any one of claims 1 to 12. Petition 870260069218, dated July 13, 2026, pp. 74 / 136

Images