CLOSED LOOP MONITORING AND CONTROL OF A CHEMICAL INJECTION SYSTEM

Abstract

An example method includes implementing a flow control device at an injection point downhole in a borehole of a wellbore operation to collect flow rate data of a fluid injected into the borehole. The method further includes collecting the flow rate data using a flow sensor disposed in the flow control device. The method further includes controlling a chemical injection system based at least in part on the flow rate data.

Description

BACKGROUND

Embodiments described herein relate generally to downhole exploration and production efforts in the resource recovery industry and more particularly to techniques for closed loop monitoring and control of a chemical injection system.

Chemical injection systems can be used to improve aspects of hydrocarbon exploration and recovery operations. For example, a chemical injection system can inject chemical(s) downhole in a borehole of a hydrocarbon exploration and recovery operation. Chemical injection systems can inhibit harmful deposition, inhibit deterioration of equipment, maintain flow capacity, avoid production rate declines, and improve production rates and equipment performance while reducing downtime risks.

In some cases, chemical injection systems provide for stimulation of hydrocarbon product. Stimulation of hydrocarbon production increases production by improving the flow of hydrocarbons into a borehole from a reservoir. Various techniques may be employed to stimulate hydrocarbon production. For example, acid stimulation may be performed, in which an acid is flowed downhole within a tubular disposed in a borehole and released into the borehole to treat the formation and stimulate fluid flow into or from the formation. After release of the acid from the tubular, hydrocarbons are received by the tubular.

SUMMARY

In one embodiment, a method is provided for closed loop monitoring and control of a chemical injection system. The method includes implementing a flow control device at an injection point downhole in a borehole of a wellbore operation to collect flow rate data of a fluid injected into the borehole. The method further includes collecting the flow rate data using a flow sensor disposed in the flow control device. The method further includes controlling a chemical injection system based at least in part on the flow rate data.

In another embodiment a system for modeling acid distribution for acid stimulation of a formation is provided. The system includes a flow control device disposed in a borehole of a wellbore operation, the flow control device including a flow sensor. The system further includes a chemical injection system to inject a fluid downhole in the borehole. The system further includes a processing system for executing computer readable instructions. The computer readable instructions control the processing system to perform operations. The operations include receiving, from the flow sensor, flow rate data indicating a flow rate of the fluid into the borehole. The operations further include controlling the chemical injection system based at least in part on the flow rate data.

In another embodiment, a method is provided for closed loop monitoring and control of a chemical injection system. The method includes injecting, using the chemical injection system, a chemical downhole in a borehole of a wellbore operation. The method includes collecting, using a flow sensor, flow rate data indicating a rate of flow of the chemical into the borehole. The method includes performing a state estimation for the chemical injection system based at least in part on the flow rate data. The method includes generating a chemical injection flow rate profile based at least in part on the flow rate data. The method includes generating a control signal to control the chemical injection system based at least in part on the chemical injection flow rate profile.

Other embodiments of the present invention implement features of the above-described method in computer systems and computer program products.

Additional technical features and benefits are realized through the techniques of the present invention. Embodiments and aspects of the invention are described in detail herein and are considered a part of the claimed subject matter. For a better understanding, refer to the detailed description and to the drawings.

BRIEF DESCRIPTION OF THE DRAWINGS

Referring now to the drawings wherein like elements are numbered alike in the several figures:

FIG. 1 depicts block diagram of a system for well production according to one or more embodiments described herein;

FIG. 2 depicts a block diagram of the surface processing unit of FIG. 1, which can be used for implementing the techniques according to one or more embodiments described herein;

FIGS. 3A and 3B depict a flow control device according to one or more embodiments described herein;

FIG. 4 depicts a flow diagram of a method for closed loop monitoring and control of a chemical injection system according to one or more embodiments described herein;

FIG. 5 depicts an example of a state estimator according to one or more embodiments described herein;

FIG. 6 depicts a block diagram of a closed loop monitoring and control system according to one or more embodiments described herein;

FIG. 7 depicts a flow diagram of a method for generating a model for closed loop monitoring and control of a chemical injection system according to one or more embodiments described herein;

FIG. 8 depicts an approach to transmitting data with respect to the hydrocarbon production system of FIG. 1 according to one or more embodiments described herein;

FIG. 9 depicts an environment for collecting and transmitting data according to one or more embodiments described herein; and

FIG. 10 depicts a block diagram of components of a machine learning training and inference system according to one or more embodiments described herein.

DETAILED DESCRIPTION

Apparatuses, systems and methods are provided for techniques for closed loop monitoring and control of a chemical injection system.

Referring to FIG. 1, an embodiment of a hydrocarbon production system 10, which operates at a wellbore operation, includes a borehole string 12 configured to be disposed in a borehole 14 that penetrates at least one earth formation 16. The borehole may be an open hole, a cased hole, or a partially cased hole. In one embodiment, the borehole string 12 is a production string that includes a tubular 18, such as a pipe (e.g., multiple pipe segments) or coiled tubing, that extends from a wellhead 20 at a surface location (e.g., at a drill site or offshore stimulation vessel). A “borehole string” as described herein may refer to any structure suitable for being lowered into a wellbore or for connecting a drill or downhole tool to the surface, and is not limited to the structure and configuration described herein. For example, the borehole string may be configured as a wireline tool, coiled tubing, a drillstring, or a logging while drilling (LWD) string.

The hydrocarbon production system 10 includes one or more assemblies 22 configured to control injection of fluid (e.g., stimulation fluid) and direct the fluid into one or more production zones in the formation 16. Each assembly 22 includes one or more flow control devices 24 (also referred to as “flow devices” or “injection devices”) configured to direct fluid from a conduit in the tubular 18 to the borehole 14. As used herein, the term “fluid” or “fluids” includes liquids, gases, hydrocarbons, multi-phase fluids, mixtures of two of more fluids, water and fluids injected from the surface, such as water or stimulation fluids. Stimulation fluids may include any suitable fluid used to reduce or eliminate an impediment to fluid production. A fluid source 26 may be coupled to the wellhead 20, and fluid from the fluid source 26 may be injected into the borehole string 12, such as by a chemical injection system (not shown).

In one embodiment, the stimulation fluid is an acid stimulation fluid. Examples of acid stimulation fluids include acids such as, but not limited to, hydrochloric acid (HCl), hydrofluoric acid, acetic acid, formic acid, sulfamic acid, chloracetic acid, carboxylic acids, organic acids and chelating agents, retarded acids, any other acid capable of dissolving the subterranean formation, and any combination thereof. Examples of chelating agents that may be suitable for use in accordance with one or more embodiments described herein include, but are not limited to, ethylenediaminetetraacetic acid (“EDTA”), glutamic acid N,N-diacetic acid (“GLDA”), and any combination thereof. Acid stimulation is useful for, e.g., removing the skin on the borehole wall that can form when a wellbore is formed in a formation, such as a carbonate formation or another suitable type of formation.

The flow control devices 24 may be any suitable structure or configuration capable of injecting or flowing fluid from the borehole string 12 and/or tubular 18 to the borehole. Examples flow control devices include flow apertures, flow input or jet valves, injection nozzles, sliding sleeves and perforations. In one embodiment, fluid is injected from the surface fluid source 26 through the tubular 18 to a sliding sleeve interface configured to provide fluid communication between the tubular 18 and a borehole annulus. The fluid can be injected into an annulus formed between the tubular 18 and the borehole wall and/or from an end of the tubular, e.g., from a coiled tubing.

According to an embodiment, the tubular 18 includes a cemented and perforated liner (not shown) and packers 42 disposed at selected locations to define isolated sections of the borehole 14 into which fluid is injected. These isolated sections are referred to as stimulation zones, each of which corresponds to a selected production zone of the formation. In each stimulation zone, at least one flow control device 24, such as one or more sliding sleeve devices, provides fluid communication between the tubular 18 and the borehole 14.

Various sensors or sensing assemblies may be disposed in and/or used by the hydrocarbon production system 10 to measure downhole parameters and conditions. For example, pressure and/or temperature sensors may be disposed at the production string at one or more locations (e.g., at or near the flow control devices 24). Other types of sensors can also be implemented. Such sensors may be configured as discrete sensors such as pressure/temperature sensors or distributed sensors. An example distributed sensor is a Distributed Temperature Sensor (DTS) assembly 28 that is disposed along a selected length of the borehole string 12. The DTS assembly 28 extends, for example, along the entire length of the string 12 between the surface and the end of the string (e.g., a toe end) or extends along selected length(s) corresponding to flow control devices 24 and/or production zones. According to an embodiment, the DTS assembly 28 is configured to measure temperature continuously or intermittently along a selected length of the string 12 and includes at least one optical fiber that extends along the string 12 (e.g., on an outside surface of the string or the tubular 18). Temperature measurements collected via the DTS assembly 28 can be used in a model to estimate fluid flow parameters in the string 12 and the borehole 14 (e.g., to estimate acid distribution in the formation 16 and/or production zones).

According to an example, one or more of the flow control devices 24 includes a flow sensor to measure flow rate data of the fluid through the one or more of the flow control devices 24. The flow sensor (see, e.g., flow sensor 310 of FIG. 3B) at the flow control device 24 provides for collecting data at the injection point of the fluid, which is further described herein.

It is understood that one or more embodiments described herein are capable of being implemented in conjunction with any suitable type of computing environment now known or later developed. In one embodiment, the DTS assembly 28, the flow control devices 24 (including the flow sensor), and/or other components are in communication with one or more processing systems, such as a surface processing unit 30 and/or a downhole electronics unit 32. The communication incorporates any of various transmission media and connections, such as wired connections, fiber optic connections, and wireless connections. The surface processing unit 30, the downhole electronics unit 32, and/or the DTS assembly 28 can include components to provide for processing, storing, and/or transmitting data collected from various sensors associated therewith.

Examples of components include, without limitation, at least one processor, storage, memory, input devices, output devices, and the like. For example, the surface processing unit 30 includes a processor 34 including a memory 36 and configured to execute software for processing measurements and generating a model as described below. As examples, one or more of the embodiments described herein can be implemented as instructions stored on a computer-readable storage medium, as hardware modules, as special-purpose hardware (e.g., application specific hardware, application specific integrated circuits (ASICs), application specific special processors (ASSPs), field programmable gate arrays (FPGAs), as embedded controllers, hardwired circuitry, etc.), or as some combination or combinations of these. According to aspects of the present disclosure, the features and functionality described herein can be a combination of hardware and programming. The programming can be processor executable instructions stored on a tangible memory, and the hardware can include the processor 34 for executing those instructions. Thus a system memory (e.g., the memory 36) can store program instructions that when executed by the processor 34 implement the features and functionality described herein.

FIG. 2 depicts a block diagram of the surface processing unit 30 of FIG. 1, which can be used for implementing the techniques described herein. In examples, the surface processing unit 30 has one or more central processing units 221a, 221b, 221c, etc. (collectively or generically referred to as processor(s) 221 and/or as processing device(s) 221). In aspects of the present disclosure, each processor 221 can include a reduced instruction set computer (RISC) microprocessor. Processors 221 are coupled to system memory (e.g., random access memory (RAM) 224) and various other components via a system bus 233. Read only memory (ROM) 222 is coupled to system bus 233 and can include a basic input/output system (BIOS), which controls certain basic functions of surface processing unit 30.

Further illustrated are an input/output (I/O) adapter 227 and a network adapter 226 coupled to system bus 233. I/O adapter 227 can be a small computer system interface (SCSI) adapter that communicates with a memory, such as a hard disk 223 and/or a tape storage device 225 or any other similar component. I/O adapter 227 and memory, such as hard disk 223 and tape storage device 225 are collectively referred to herein as mass storage 234. Operating system 240 for execution on the surface processing unit 30 can be stored in mass storage 234. The network adapter 226 interconnects system bus 233 with an outside network 236 enabling the surface processing unit 30 to communicate with other systems.

A display 235 (e.g., a display monitor) is connected to system bus 233 by display adaptor 232, which can include a graphics adapter to improve the performance of graphics intensive applications and a video controller. In one aspect of the present disclosure, adapters 226, 227, and/or 232 can be connected to one or more I/O busses that are connected to system bus 233 via an intermediate bus bridge (not shown). Suitable I/O buses for connecting peripheral devices such as hard disk controllers, network adapters, and graphics adapters typically include common protocols, such as the Peripheral Component Interconnect (PCI). Additional input/output devices are shown as connected to system bus 233 via user interface adapter 228 and display adapter 232. A keyboard 229, mouse 230, and speaker 231 can be interconnected to system bus 233 via user interface adapter 228, which can include, for example, a Super I/O chip integrating multiple device adapters into a single integrated circuit.

In some aspects of the present disclosure, the surface processing unit 30 includes a graphics processing unit 237. Graphics processing unit 237 is a specialized electronic circuit designed to manipulate and alter memory to accelerate the creation of images in a frame buffer intended for output to a display. In general, graphics processing unit 237 is very efficient at manipulating computer graphics and image processing and has a highly parallel structure that makes it more effective than general-purpose CPUs for algorithms where processing of large blocks of data is done in parallel.

Thus, as configured herein, the surface processing unit 30 includes processing capability in the form of processors 221, storage capability including system memory (e.g., RAM 224 and/or mass storage 234), input means such as keyboard 229 and mouse 230, and output capability including speaker 231 and display 235. In some aspects of the present disclosure, a portion of system memory (e.g., RAM 224 and mass storage 234) collectively store the operating system 240 to coordinate the functions of the various components shown in the surface processing unit 30.

One or more embodiments described herein provides for closed loop monitoring and control of a chemical injection system. For example, an embodiment includes implementing a flow control device (e.g., the one or more flow control devices 24) at an injection point downhole in a wellbore of an to collect flow rate data. The flow rate data indicates the rate at which fluid (e.g., chemicals), which is supplied via a chemical injection section, flows through the flow control device and into the wellbore. The chemical injection system is then controlled based at least in part on the flow rate data. For example, if the flow rate data indicates that the flow through the flow control device is less than an expected flow rate, the chemical injection system can be controlled to cause the flow rate to increase. Conversely, as another example, if the flow rate data indicates that the flow through the flow control device is greater than an expected flow rate, the chemical injection system can be controlled to cause the flow rate to decrease.

Conventional chemical injection systems provide fluid (e.g., chemicals) from the surface at a certain rate to the flow control devices downhole. However, such conventional chemical injection systems cannot confirm that the fluid actually injected at the flow control devices is the same amount or rate that is expected. For example, leakage can occur, which can cause the expected amount or rate of fluid (e.g., chemicals) not to be realized.

In an effort to cure this and other shortcomings, one or more embodiments monitor flow rate at flow control devices and then control the chemical injection system accordingly. For example, by installing flow sensors downhole at the flow control devices, the actual injection rate can be measured in real time (or near real time). As described herein, according to an example, one or more of the flow control devices 24 includes a flow sensor to measure flow rate data of the fluid through the one or more of the flow control devices 24. The flow sensor (see, e.g., flow sensor 310 of FIG. 3B) at the flow control device 24 provides for collecting data at the injection point of the fluid. Any leakage in the system can be detected by the flow sensor via the flow rate data transmitted to the surface (e.g., to the surface processing unit 30). According to an embodiment, a machine learning model can be used to identify leakage based on the flow rate data. The chemical injection system can then be controlled to adjust flow rate based on the flow rate data.

FIG. 3A is a schematic illustration of a flow control device 300, which is an example of one or more of the flow control devices 24 of FIG. 1. A chemical injection system pumps fluids (e.g., chemicals) using a surface pump to the flow control device 300, which is disposed in the borehole 14 as shown in FIG. 1. The chemical is injected through the flow control device 300 into the borehole 14.

FIG. 3B is a schematic illustration of an internal portion of the flow control device 300. In this example, the flow control device 300 includes a flow sensor 310, which measures fluid (e.g., chemical) flow through the flow control device 300 and into the borehole 14.

FIG. 4 depicts a flow diagram of a method for closed loop monitoring and control of a chemical injection system according to one or more embodiments described herein. The method 400 can be performed by any suitable processing system downhole or on surface (e.g., the surface processing unit 30, a downhole electronics unit 32, a cloud computing node of a cloud computing environment, etc.), any suitable processing device (e.g., the processor 34, one of the processors 21), and/or combinations thereof or another suitable system or device.

At block 402, a chemical injection system (not shown) injects a fluid (e.g., a chemical) downhole in the borehole 14. For example, the chemical injection system injects a chemical downhole through a control line through one or more of the flow control devices 24 (e.g., the flow control device 300 of FIG. 3) to treat organic and/or inorganic scale in the borehole.

At block 404, a flow sensor (e.g., the flow sensor 310 of the flow control device 300) collects flow rate data at one or more of the flow control devices 24. For example, the flow sensor 310 can measure differential pressure and acoustic emission energy at the flow control device 300. The flow rate data can be transmitted to the surface (e.g., to the surface processing unit 30) using any suitable wired and/or wireless transmission technique, such as mud pulse telemetry.

At block 406, a processing system (e.g., the surface processing unit 30) performs a state estimation for the chemical injection system using a state estimator. For example, a state estimator receives data from multiple flow sensors (e.g., multiple of the flow sensors 310), which can be in the same or multiple flow control devices. The state estimator can perform data fusion and assimilation on the flow rate data from the multiple sensors.

An example of a state estimator 500 is shown in FIG. 5 according to one or more embodiments described herein. The state estimator 500 uses data, received from data sources 502, such as from a database management system 502a, data warehouses 502b, data streams 502c, and/or other data 502d such as well logs, seismic data, and/or the like, including combinations and/or multiples thereof. An example of one of the data sources 502 is the flow sensor 310. Another example of the data sources 502 is multiple flow sensors. The flow sensor 310 (or multiple flow sensors) can provide flow rate data to the state estimator 500 as a data stream for the data streams 502c, for example, or can store the flow rate data in a data warehouse of the data warehouses 502b or other suitable data storage location.

The state estimator 500 performs data management 504 using one or more of the data sources 502. The data management can include transforming the data 504a, performing pre- and/or post-processing of the data 504b, filtering the data 504c, aggregating the data 504d, storing the data 504e, such as in distributed data storage, and/or the like, including combinations and/or multiples thereof.

At block 506, the state estimator 500 performs modeling, which can be performed, for example, using a predictive model markup language (PMML) or another suitable technique. Modeling at block 506 can include a proxy-based model and state estimation 506a, model validation 506b, and model scoring, calibration, and updating 506c. The proxy-based model and state estimation 506a uses the output from the data management to build a model, which estimates states for the chemical injection system. That is, based on measurements collected at the flow control devices 24 (e.g., the flow rate data), the modeling at block 506 generates a model to estimate a state of the chemical injection system. The model can be validated using, for example, validation data or other validation techniques. The model can then be scored, calibrated, and updated, such as when new flow rate data is received.

The state estimator 500 can also provide for visualization and analytics results 508. For example, the state estimator can provide one or more of the following: multiphase flow estimation and allocation; events detection, diagnosis and prognosis; system future production forecasting; set-points optimization for control system; water and gas breakthrough or coning detection; water and gas cross flow detection; visualization of production key performance indicators (KPIs); data-driven petrophysical properties estimation; well production management; reservoir production management, and/or the like, including combinations and/or multiples thereof.

With continued reference to FIG. 4, at block 408, the processing system generates a chemical injection flow rate profile. For example, the processing system generates optimal chemical injection flow rates based on the current state estimation (from block 406) and a desired outcome/goal (e.g., a production rate).

At block 410, the processing system generates control signals to control the flow rate through the flow control devices 24 and to control the chemical injection system to provide fluid (e.g., chemicals) into the borehole 14 at a desired rate. According to one or more embodiments described herein, the processing system can utilize artificial intelligence to control and automate the chemical injection system. Machine learning is described further herein with reference to FIG. 10.

Additional processes also may be included, and it should be understood that the process depicted in FIG. 4 represents an illustration, and that other processes may be added or existing processes may be removed, modified, or rearranged without departing from the scope of the present disclosure.

FIG. 6 depicts a block diagram of a closed loop monitoring and control system 600 according to one or more embodiments described herein. At block 602, a surveillance system (e.g., the surface processing unit 30 in combination with one or more flow sensors 310) send flow rate data (e.g., multiple sensor data) to the state estimator 604, which is an example of the state estimator 500 of FIG. 5. The state estimator 604 generates estimated system state calibrated parameters and sends these to an optimization module 606. The optimization module includes a state prediction block 608 and a production optimization block 610. The state prediction block 608 generates indications of system performance, such as flow rates and KPIs, and the production optimization block 610 defines well set points. The optimization module 606 can repeat the cycle between block 608, 610 iteratively to run iterations to reach an optimal flow vector, which is then transmitted to the surveillance system at block 602 as optimal well setpoints and/or predicted system KPIs.

The closed loop monitoring and control system 600 can be implemented as the system 620. The system 620 receives desired flow rate as a system input to system block 622. The system block 622 outputs a measured downhole flow rate at an injection point as determined by a flow sensor (e.g., the flow sensor 310), which is fed into state estimator 624, which is an example of the state estimator 500. The desired flow rate is also input into a system model 626 as a system input. The system model 626 calculates an expected downhole flow rate, which is a desired amount of flow at the injection point. The state estimator 624 receives the expected flow rate. The state estimator 624 compares the measured downhole flow rate and the expected downhole flow rate to determine whether the rates match or differ. The results are input into an adaptive optimization based identifier 628, which determines, for example, a production rate based on the difference from the state estimator 624. The adaptive optimization based identifier 628 can identify model parameters to adjust the model at the system block 622 as shown.

FIG. 7 depicts a flow diagram of a method 700 for generating a model for closed loop monitoring and control of a chemical injection system according to one or more embodiments described herein. The method 400 can be performed by any suitable processing system downhole or on surface (e.g., the surface processing unit 30, a downhole electronics unit 32, a cloud computing node of a cloud computing environment, etc.), any suitable processing device (e.g., the processor 34, one of the processors 21), and/or combinations thereof or another suitable system or device.

At block 702, design of experiment (DOE) is performed. Design of experiment includes establishing input variables (e.g., flow rate) and output variables (e.g., change in flow control device, change in chemical injection system, and/or the like, including combinations and/or multiples thereof). At block 704, DOE sampling of a reservoir/well of interest is performed to collect data (e.g., flow rate data) about the reservoir/well. At block 706, the processing system calculates responses for the system, such as flow rates based on the flow rate data. At block 708, data preprocessing can be performed, such as correlating the inputs/outputs.

Optimization 709 begins at block 710 where the processing system applies the model from FIGS. 5 and/or 6 to perform a dynamic data driven prediction and calibration. This can be accomplished, for example, using machine learning techniques as describe herein, such as multiple input multiple output, artificial neural networks, support vector machines, decision trees, and/or the like, including combinations and/or multiples thereof. The model at block 710 can consider a well head choak position from block 712, electric submersible pump (ESP) parameters from block 714, well and reservoir parameters from block 716, and/or observed response and load disturbance 718. The model at block 710 generates predicted and observed ESP performance parameters, which are input into a decision block 720. At decision block 720, it is determined whether the predicted and observed responses are satisfactory (e.g., whether the responses satisfy a threshold) based on optimization objectives and constraints 722. If so (“yes” at decision block 720), the method 700 proceeds to block 724, where supervisory system tests and validation can be performed at decision block 726. If validation is successful (“yes” at decision block 726), optimal flow estimation is performed at block 728 to determine the optimal fluid flow for the chemical injection system. If validation is unsuccessful (“no at decision block 726), the method 700 returns to the optimization 709, which can then be performed again.

Returning to decision block 720, if it is determined that the predicted and observed responses are not satisfactory (“no” at decision block 720), the parameters at block 712, 714, 716 are updated at block 730, and the model is updated at block 710. The optimization 709 can be performed again.

Additional processes also may be included, and it should be understood that the process depicted in FIG. 7 represents an illustration, and that other processes may be added or existing processes may be removed, modified, or rearranged without departing from the scope of the present disclosure.

FIG. 8 depicts an approach to transmitting data with respect to the hydrocarbon production system 10 according to one or more embodiments described herein. For example, a collection layer 802, the approach to transmitting data includes collecting data at flow sensors (e.g., flow sensor 310 of FIG. 3B) within flow control devices (e.g., flow control devices 24) downhole in the borehole 14. The data are transmitted to a layer 804, which collects and aggregates the data. The data are then transmitted to a layer 806, which performs data management (e.g., data storage, in-memory data processing, and/or the like, including combinations and/or multiples thereof). The data are then transmitted to a layer 808, where data analytics are performed (e.g., data processing, data mining, and/or the like, including combinations and/or multiples thereof). The data are then transmitted to a layer 810, which provides data-driven decision support (e.g., simulation optimization; ESP, well, and reservoir management; and/or the like, including combinations and/or multiples thereof). The data are then transmitted to a layer 812, where control and visualization can occur, such as intelligent control of a chemical injection system 814, real-time publishing of the data to one or more user devices 816 or the web 818, and/or the like, including combinations and/or multiples thereof.

FIG. 9 depicts an environment for collecting and transmitting data according to one or more embodiments described herein. In this example, data collected by devices 902 (e.g., the flow sensor 310) at a well site 900. Then, the data is transmitted from the well site 900 to a data hosting facility 910 (e.g., a cloud computing environment) as shown. The data are then available to access from a data access and control location 920 (e.g., a user device). Control commands (e.g., for controlling the chemical injection system) can be sent to the well site 900 from the data access and control location 920 via the data hosting facility 910 as shown. According to one or more embodiments described herein, data can be visualized and/or made available to a third-party as shown.

As described herein, one or more embodiments described herein can utilize machine learning techniques to perform tasks in conjunction with closed loop monitoring and control of a chemical injection system. More specifically, one or more embodiments described herein can incorporate and utilize rule-based decision making and artificial intelligence (AI) reasoning to accomplish the various operations described herein, such as generating control signals to control the flow rate of fluid (e.g., chemicals) through the flow control deices 24 and/or to control the chemical injection system to provide fluid (e.g., chemicals) into the borehole 14 at a desired rate (see, e.g., block 410 of FIG. 4). The phrase “machine learning” broadly describes a function of electronic systems that learn from data. A machine learning system, engine, or module can include a trainable machine learning algorithm that can be trained, such as in an external cloud environment, to learn functional relationships between inputs and outputs, and the resulting model (sometimes referred to as a “trained neural network,” “trained model,” and/or “trained machine learning model”) can be used in conjunction with closed loop monitoring and control of a chemical injection system, for example. In one or more embodiments, machine learning functionality can be implemented using an artificial neural network (ANN) having the capability to be trained to perform a function. In machine learning and cognitive science, ANNs are a family of statistical learning models inspired by the biological neural networks of animals, and in particular the brain. ANNs can be used to estimate or approximate systems and functions that depend on a large number of inputs. Convolutional neural networks (CNN) are a class of deep, feed-forward ANNs that are particularly useful at tasks such as, but not limited to analyzing visual imagery and natural language processing (NLP).

ANNs can be embodied as so-called “neuromorphic” systems of interconnected processor elements that act as simulated “neurons” and exchange “messages” between each other in the form of electronic signals. Similar to the so-called “plasticity” of synaptic neurotransmitter connections that carry messages between biological neurons, the connections in ANNs that carry electronic messages between simulated neurons are provided with numeric weights that correspond to the strength or weakness of a given connection. The weights can be adjusted and tuned based on experience, making ANNs adaptive to inputs and capable of learning. For example, an ANN for handwriting recognition is defined by a set of input neurons that can be activated by the pixels of an input image. After being weighted and transformed by a function determined by the network's designer, the activation of these input neurons are then passed to other downstream neurons, which are often referred to as “hidden” neurons. This process is repeated until an output neuron is activated. The activated output neuron determines which character was input. It should be appreciated that these same techniques can be applied in the case of generating control signals for closed loop monitoring and control of a chemical injection system as described herein.

Systems for training and using a machine learning model are now described in more detail with reference to FIG. 10. Particularly, FIG. 10 depicts a block diagram of components of a machine learning training and inference system 1000 according to one or more embodiments described herein. The system 1000 performs training 1002 and inference 1004. During training 1002, a training engine 1016 trains a model (e.g., the trained model 1018) to perform a task, such as to generate control signals for closed loop monitoring and control of a chemical injection system. Inference 1004 is the process of implementing the trained model 1018 to perform the task, such as to generate control signals for closed loop monitoring and control of a chemical injection system, in the context of a larger system (e.g., a system 1026). All or a portion of the system 1000 shown in FIG. 10 can be implemented, for example by all or a subset of the surface processing unit 30.

The training 1002 begins with training data 1012, which may be structured or unstructured data. According to one or more embodiments described herein, the training data 1012 includes flow rates. The training engine 1016 receives the training data 1012 and a model form 1014. The model form 1014 represents a base model that is untrained. The model form 1014 can have preset weights and biases, which can be adjusted during training. It should be appreciated that the model form 10110 can be selected from many different model forms depending on the task to be performed. For example, where the training 1002 is to train a model to perform image classification, the model form 1014 may be a model form of a CNN. The training 1002 can be supervised learning, semi-supervised learning, unsupervised learning, reinforcement learning, and/or the like, including combinations and/or multiples thereof. For example, supervised learning can be used to train a machine learning model to classify an object of interest in an image. To do this, the training data 1012 includes labeled images, including images of the object of interest with associated labels (ground truth) and other images that do not include the object of interest with associated labels. In this example, the training engine 1016 takes as input a training image from the training data 1012, makes a prediction for classifying the image, and compares the prediction to the known label. The training engine 1016 then adjusts weights and/or biases of the model based on results of the comparison, such as by using backpropagation. The training 1002 may be performed multiple times (referred to as “epochs”) until a suitable model is trained (e.g., the trained model 1018).

According to one or more embodiments described herein, the training 1002 can apply supervised learning to train a decision tree, which can be used to generate control signals for closed loop monitoring and control of a chemical injection system. There are generally two types of decision trees: classification trees and regression trees. Classification trees take as input a discrete set of values, and leaves of the tree represent classes labels with branches representing conjunctions that result in those class labels. Regression trees take as input a target variable as a continuous set of values. Decision trees are useful to predict the value of a target variable based on input variables. A decision tree can be trained and used to predict a target flow rate for a chemical injection system based on flow rate data from multiple flow control devices. It should be appreciated that other types of machine learning can also be applied to the present techniques.

Once trained, the trained model 1018 can be used to perform inference 1004 to perform a task, such as to generate control signals for closed loop monitoring and control of a chemical injection system. The inference engine 1020 applies the trained model 1018 to new data 1022 (e.g., real-world, non-training data). For example, if the trained model 1018 is trained to classify images of a particular object, such as a chair, the new data 1022 can be an image of a chair that was not part of the training data 1012. In this way, the new data 1022 represents data to which the model 1018 has not been exposed. The inference engine 1020 makes a prediction 1024 (e.g., a classification of an object in an image of the new data 1022) and passes the prediction 1024 to the system 1026 (e.g., the surface processing unit 30). The system 1026 can, based on the prediction 1024, taken an action, perform an operation, perform an analysis, and/or the like, including combinations and/or multiples thereof. In some embodiments, the system 1026 can add to and/or modify the new data 1022 based on the prediction 1024.

In accordance with one or more embodiments, the predictions 1024 generated by the inference engine 1020 are periodically monitored and verified to ensure that the inference engine 1020 is operating as expected. Based on the verification, additional training 1002 may occur using the trained model 1018 as the starting point. The additional training 1002 may include all or a subset of the original training data 1012 and/or new training data 1012. In accordance with one or more embodiments, the training 1002 includes updating the trained model 1018 to account for changes in expected input data.

Example embodiments of the disclosure include or yield various technical features, technical effects, and/or improvements to technology. Example embodiments of the disclosure provide technical solutions for closed loop monitoring and control of a chemical injection system. The techniques described herein represent an improvement to conventional chemical injection systems. Specifically, one or more embodiments described herein collect flow rate data downhole at an injection point (or multiple injection points) and use that data to determine how to control the chemical injection system (e.g., increase/decrease flow rate of fluid (e.g., chemicals)). Additionally or alternatively, one or more embodiments described herein apply machine learning techniques to determine how to control chemical injection systems. Each of these improvements increases hydrocarbon recovery from a hydrocarbon reservoir compared to conventional techniques by more precisely controlling the flow of fluid (e.g., chemicals) through injection points (e.g., flow control devices). Moreover, chemical use can be reduced because one or more embodiments apply a desired flow rate more accurately, thus avoiding excessive flow.

Set forth below are some embodiments of the foregoing disclosure:

Embodiment 1: A method comprises implementing a flow control device at an injection point downhole in a borehole of a wellbore operation to collect flow rate data of a fluid injected into the borehole; collecting the flow rate data using a flow sensor disposed in the flow control device; and controlling a chemical injection system based at least in part on the flow rate data.

Embodiment 2: A method according to any prior embodiment, wherein controlling the chemical injection system comprises determining a difference between an expected flow rate and an actual flow rate, the expected flow rate being a flow rate provided by the chemical injection system, and the actual flow rate being a flow rate at an injection point as determined by the flow rate data.

Embodiment 3: A method according to any prior embodiment, wherein controlling the chemical injection system comprises adjusting the flow rate provided by the chemical injection system responsive the difference satisfying a threshold.

Embodiment 4: A method according to any prior embodiment, wherein the chemical injection system is controlled based at least in part on a machine learning model.

Embodiment 5: A method according to any prior embodiment, wherein the machine learning model is a decision tree.

Embodiment 6: A method according to any prior embodiment, wherein the chemical injection system is controlled based at least in part on a state estimator.

Embodiment 7: A method according to any prior embodiment, further comprising injecting, using the chemical injection system, the fluid downhole in the borehole.

Embodiment 8: A system comprising a flow control device disposed in a borehole of a wellbore operation, the flow control device comprising a flow sensor; a chemical injection system to inject a fluid downhole in the borehole; and a processing system for executing computer readable instructions, the computer readable instructions controlling the processing system to perform operations comprising: receiving, from the flow sensor, flow rate data indicating a flow rate of the fluid into the borehole; and controlling the chemical injection system based at least in part on the flow rate data.

Embodiment 9: A system according to any prior embodiment, wherein controlling the chemical injection system comprises determining a difference between an expected flow rate and an actual flow rate, the expected flow rate being a flow rate provided by the chemical injection system, and the actual flow rate being a flow rate at an injection point as determined by the flow rate data.

Embodiment 10: A system according to any prior embodiment, wherein controlling the chemical injection system comprises adjusting the flow rate provided by the chemical injection system responsive the difference satisfying a threshold.

Embodiment 11: A system according to any prior embodiment, wherein the chemical injection system is controlled based at least in part on a machine learning model.

Embodiment 12: A system according to any prior embodiment, wherein the machine learning model is a decision tree.

Embodiment 13: A system according to any prior embodiment, wherein the chemical injection system is controlled based at least in part on a state estimator.

Embodiment 14: A system according to any prior embodiment, wherein the flow control device is a first flow control device, wherein the flow sensor is a first flow sensor, and wherein the flow rate data is first flow rate data, the system further comprising: a second flow control device disposed in the borehole, the second flow control device comprising a second flow sensor to collect second flow rate data.

Embodiment 15: A system according to any prior embodiment, wherein controlling the chemical injection system is system based at least in part on the first flow rate data and the second flow rate data.

Embodiment 16: A method comprises injecting, using a chemical injection system, a chemical downhole in a borehole of a wellbore operation; collecting, using a flow sensor, flow rate data indicating a rate of flow of the chemical into the borehole; performing a state estimation for the chemical injection system based at least in part on the flow rate data; generating a chemical injection flow rate profile based at least in part on the flow rate data; and generating a control signal to control the chemical injection system based at least in part on the chemical injection flow rate profile.

Embodiment 17: A system according to any prior embodiment, wherein the control signal is determined using a machine learning model, wherein the machine learning model is a decision tree.

Embodiment 18: A system according to any prior embodiment, wherein the flow sensor is disposed in a flow control device.

Embodiment 19: A system according to any prior embodiment, wherein the control signal is a signal to increase the rate of flow of the chemical into the borehole.

Embodiment 20: A system according to any prior embodiment, wherein the control signal is a signal to decrease the rate of flow of the chemical into the borehole.

The use of the terms “a” and “an” and “the” and similar referents in the context of describing the present disclosure (especially in the context of the following claims) are to be construed to cover both the singular and the plural, unless otherwise indicated herein or clearly contradicted by context. Further, it should further be noted that the terms “first,” “second,” and the like herein do not denote any order, quantity, or importance, but rather are used to distinguish one element from another. The modifier “about” used in connection with a quantity is inclusive of the stated value and has the meaning dictated by the context (e.g., it includes the degree of error associated with measurement of the particular quantity).

The teachings of the present disclosure can be used in a variety of well operations. These operations can involve using one or more treatment agents to treat a formation, the fluids resident in a formation, a wellbore, and/or equipment in the wellbore, such as production tubing. The treatment agents can be in the form of liquids, gases, solids, semi-solids, and mixtures thereof. Illustrative treatment agents include, but are not limited to, fracturing fluids, acids, steam, water, brine, anti-corrosion agents, cement, permeability modifiers, drilling muds, emulsifiers, demulsifiers, tracers, flow improvers etc. Illustrative well operations include, but are not limited to, hydraulic fracturing, stimulation, tracer injection, cleaning, acidizing, steam injection, water flooding, cementing, etc.

While the present disclosure has been described with reference to an embodiment or embodiments, it will be understood by those skilled in the art that various changes can be made and equivalents can be substituted for elements thereof without departing from the scope of the present disclosure. In addition, many modifications can be made to adapt a particular situation or material to the teachings of the present disclosure without departing from the essential scope thereof. Therefore, it is intended that the present disclosure not be limited to the particular embodiment disclosed as the best mode contemplated for carrying out this present disclosure, but that the present disclosure will include all embodiments falling within the scope of the claims. Also, in the drawings and the description, there have been disclosed embodiments of the present disclosure and, although specific terms can have been employed, they are unless otherwise stated used in a generic and descriptive sense only and not for purposes of limitation, the scope of the present disclosure therefore not being so limited.

Claims

1. A method comprising:

implementing a flow control device at an injection point downhole in a borehole of a wellbore operation to collect flow rate data of a fluid injected into the borehole;

collecting the flow rate data using a flow sensor disposed in the flow control device; and

controlling a chemical injection system based at least in part on the flow rate data.

2. The method of claim 1, wherein controlling the chemical injection system comprises determining a difference between an expected flow rate and an actual flow rate, the expected flow rate being a flow rate provided by the chemical injection system, and the actual flow rate being a flow rate at an injection point as determined by the flow rate data.

3. The method of claim 2, wherein controlling the chemical injection system comprises adjusting the flow rate provided by the chemical injection system responsive the difference satisfying a threshold.

4. The method of claim 1, wherein the chemical injection system is controlled based at least in part on a machine learning model.

5. The method of claim 4, wherein the machine learning model is a decision tree.

6. The method of claim 1, wherein the chemical injection system is controlled based at least in part on a state estimator.

7. The method of claim 1, further comprising injecting, using the chemical injection system, the fluid downhole in the borehole.

8. A system comprising:

a flow control device disposed in a borehole of a wellbore operation, the flow control device comprising a flow sensor;

a chemical injection system to inject a fluid downhole in the borehole; and

a processing system for executing computer readable instructions, the computer readable instructions controlling the processing system to perform operations comprising: receiving, from the flow sensor, flow rate data indicating a flow rate of the fluid into the borehole; and controlling the chemical injection system based at least in part on the flow rate data.

9. The system of claim 8, wherein controlling the chemical injection system comprises determining a difference between an expected flow rate and an actual flow rate, the expected flow rate being a flow rate provided by the chemical injection system, and the actual flow rate being a flow rate at an injection point as determined by the flow rate data.

10. The system of claim 9, wherein controlling the chemical injection system comprises adjusting the flow rate provided by the chemical injection system responsive the difference satisfying a threshold.

11. The system of claim 8, wherein the chemical injection system is controlled based at least in part on a machine learning model.

12. The system of claim 11, wherein the machine learning model is a decision tree.

13. The system of claim 8, wherein the chemical injection system is controlled based at least in part on a state estimator.

14. The system of claim 8, wherein the flow control device is a first flow control device, wherein the flow sensor is a first flow sensor, and wherein the flow rate data is first flow rate data, the system further comprising:

a second flow control device disposed in the borehole, the second flow control device comprising a second flow sensor to collect second flow rate data.

15. The system of claim 14, wherein controlling the chemical injection system is system based at least in part on the first flow rate data and the second flow rate data.

16. A method comprising:

injecting, using a chemical injection system, a chemical downhole in a borehole of a wellbore operation;

collecting, using a flow sensor, flow rate data indicating a rate of flow of the chemical into the borehole;

performing a state estimation for the chemical injection system based at least in part on the flow rate data;

generating a chemical injection flow rate profile based at least in part on the flow rate data; and

generating a control signal to control the chemical injection system based at least in part on the chemical injection flow rate profile.

17. The method of claim 16, wherein the control signal is determined using a machine learning model, wherein the machine learning model is a decision tree.

18. The method of claim 16, wherein the flow sensor is disposed in a flow control device.

19. The method of claim 16, wherein the control signal is a signal to increase the rate of flow of the chemical into the borehole.

20. The method of claim 16, wherein the control signal is a signal to decrease the rate of flow of the chemical into the borehole.