Methods, systems, and controllers for determining an amount of drag reduction agent

Abstract

Embodiments of systems and methods to determine an amount of drag reduction agent (DRA) to inject into pipeline are disclosed. In an embodiment, such a method may include determining parameters to include one or more of (a) properties of hydrocarbon flowing through the section of pipeline, (b) a current amount and properties of DRA injected into section of the pipeline, or (c) one or more of a speed or power of a pump for the section of pipeline. The method may include determining an adjusted amount of DRA to inject into the section of the pipeline based on application of the parameters to a trained machine learning model. The method may include injecting the adjusted amount of DRA into the section of the pipeline.

Description

METHODS, SYSTEMS, AND CONTROLLERS FOR DETERMINING AN AMOUNT OF DRAG REDUCTION AGENTCROSS-REFERENCE TO RELATED APPLICATIONS

[0001] This application claims priority to and the benefit of U.S. Provisional Patent Application No. 63 / 631,053, filed April 8, 2024, which is hereby incorporated by reference in its entirety.FIELD OF DISCLOSURE

[0002] The disclosure herein relates to systems and methods for determining an amount of drag reduction agent to and, particularly, for determining and adjusting an amount of drag reduction agent for a selected section of pipeline.BACKGROUND

[0003] Different types of hydrocarbon liquids, such as petroleum and renewable liquid products (e.g., such as crude oil), are often transported via pipeline to a refinery and / or other end-use locations. To reduce the viscosity of heavy crude and maximize capacity, drag reduction agent may be injected into each section of the pipeline to increase the flow rate of the hydrocarbon liquid, reduce the time to the refinery, and to decrease pump power utilized. The amount of drag reduction agent to be added to a pipeline is determined based on rudimentary equations that simply consider viscosity of the hydrocarbon liquids. Additionally, some operators may simply inject large quantities of drag reduction agent without consideration of any factor. Further, each section of the pipeline may include different properties or parameters. Thus, the amount of DRA injected at any section of the pipeline may be an underestimation or an overestimation.SUMMARY

[0004] Thus, in view of the foregoing, Applicant has recognized these problems and others in the art, and has recognized a need for enhanced systems and methods for determining an amount of drag reduction agent (DRA) to inject at one or more sections of pipeline. Particularly, the present disclosure relates to systems and methods for determining an amount of DRA to inject into one or more sections of pipeline and, particularly, for determining and adjusting one or more of an amount of DRA or a drag reduction amount or percentage for a selected section of pipeline using trained machine learning models.

[0005] The disclosure herein provides embodiments of systems and methods for determining and adjusting an amount of DRA to inject into each section of pipeline. In an embodiment, the systems and methods may include a computing device, apparatus and / or controller to obtain various data points and / or parameters to train a machine learning model and to apply to a trained machine learning model. The system may include, for example, for each section of a pipeline, a pump, a DRA injection device, a DRA tank or container, and / or a plurality of sensors and / or analyzers. In an example, each section of the pipeline may include a starting location, defined by a pump or pumping station, and an end location, defined by the beginning of a subsequent pump or pumping station. The DRA injection device and DRA tank or container may be positioned proximate the pump or pumping station. In other embodiments, each section of the pipeline may include additional DRA injection devices and DRA tanks or containers.

[0006] As a pipeline operation begins, hydrocarbon may begin flowing from a source to a first pump or pump station. Prior to such a pipeline operation, the systems and methods may include applying one or more parameters to a trained machine learning model. Such an application to the trained machine learning models may determine a value indicative of how much DRA should be injected into a section of the pipeline and / or an amount of drag reduction for the pipeline operation at the section of the pipeline. Such outputs may be determined based on the parameters applied to the trained machine learning model and / or based on an input indicative of the type of output desired. Upon output of such values, the systems and methods may adjust the amount of DRA to be injected or being injected into the section of the pipeline.

[0007] The embodiments described above may be iterative for each section of pipeline. For example, at the beginning of a pipeline operation, the obtained parameters may be applied to the trained machine learning model for each section of the pipeline. As the pipeline operation continues, parameters measured or obtained during the pipeline operation may be utilized to adjust the DRA amount in real-time (for example, based on application of the parameters to the trained machine learning model). Thus, adjustments may occur in realtime over the course of a pipeline operation.

[0008] In yet other embodiments, the system may include a plurality of controllers. Each one of the plurality of controllers may be positioned at each one of the sections of pipeline. Further, each controller may include the trained machine learning model. The controller may store selected parameters specific to a corresponding section of the pipeline. In other embodiments, the controller may store a selected trained machine learning model that hasbeen trained using parameters specific to that corresponding section of pipeline. In yet another embodiment, a supervisory controller or computing device may receive parameters for each section and apply those parameters to the trained machine learning model.

[0009] In an embodiment, the parameters noted above may include various measurements, properties, and / or indicators obtained in real-time and / or that are static or change less frequent rates. For example, the parameters may include pipeline diameter for each section of the pipeline, the direction that the pipeline runs (for example, at a downward angle, upward angle, or horizontally), the location of the pipeline (for example, underground or aboveground), any known internal anomalies of the pipeline (for example, rough sections of the pipeline), properties hydrocarbon for a pipeline operation (for example, a grade or grade code of the hydrocarbon and / or a specification indicating various parameters of the hydrocarbon), a viscosity of the hydrocarbon, an amount of the hydrocarbon, a flow rate of the hydrocarbon, a power or speed setting for each pump, an amount of DRA to be injected into a section of the pipeline, properties of DRA to be utilized, and / or temperature within various locations of each section of the pipeline, among other parameters. In an embodiment, to obtain an amount of DRA to inject at a section, the system may obtain a desired drag reduction amount and hydrocarbon grade code. In other embodiments, additional factors or parameters may be utilized.

[0010] As noted, the systems and methods may utilize a trained machine learning model. The systems and methods may train the trained machine learning model. To begin the training process, the systems and methods may first determine or generate a training data set. The systems and methods may include separating the historical data set into distinct data sets based on a distinct hydrocarbon flowing through the pipeline at a selected time. Each of the distinct data sets may include one or more of the parameters described above. Once a selected threshold amount of distinct data sets have been determined, the distinct data sets may be split into a training set and a testing set. The system may train the machine learning model using the training set and then test the trained machine learning model with the testing set. If the output of the trained machine learning model falls below a second selected threshold, the system may retrain the trained machine learning model with a different portion of the distinct data sets and / or with additional historical data. Once the trained machine learning model meets the second selected threshold, the system may utilize the trained machine learning model during pipeline operations for each section of the pipeline. In an embodiment, the trained machine learning model, as noted, may correspond to a section of the pipeline. In another embodiment, the trained machine learning modelmay not be pipeline section specific. In other words, the trained machine learning model may be utilized for any section of the pipeline.

[0011] By utilizing the trained machine learning model, the systems and methods described herein may determine a specific amount of DRA to be utilized in each section of pipeline during a pipeline operation. Further, the systems and methods may also determine an amount of or a percentage of drag reduction for each section of pipeline. Further, these determinations may occur prior to initialization of the pipeline operation and during the pipeline operation. Thus, as various parameters change over time (for example, hydrocarbon viscosity and / or DRA build up from previous sections of pipeline, among other factors), the systems and methods may continue to adjust DRA amount to ensure balance between injected DRA and pump power.

[0012] Thus, rather than utilizing a random amount of DRA for any section of the pipeline, the systems and methods described herein enable precise amounts of DRA (for example, at parts per million (PPM) in relation to the amount of hydrocarbon flowing through the pipeline) to be utilized from initialization to the end of the pipeline operation. Further, the amount of DRA utilized may be adjusted based on, for example pump power or speed, to ensure that the amount DRA utilized is at a minimum in relation to pump power or speed. In another embodiment, if any pumps used at any section of the pipeline are variable frequency drive (VFD) or variable speed drive (VSD) pumps, then the power utilized by a pump may be adjusted to minimize power used in relation to use of DRA. Additionally, pumps at one or more sections of the pipeline may be shut-off or shut-down to limit power use, particularly fixed speed or frequency pumps. Further, the trained machine learning model may be trained such that minimal input to the trained machine learning model may determine the precise amount of DRA for selected grades or grade codes of hydrocarbons and / or for specific or selected sections of pipeline (based on, for example, selected parameters specific to a section of pipeline).

[0013] Accordingly, an embodiment of the disclosure is directed to a method to determine an amount of DRA to inject into one or more sections of pipeline. The method may include generating a training data set based on a plurality of data sets that each correspond to a distinct hydrocarbon flow for a selected section of the one or more sections of the pipeline. The method may include training a machine learning model with the training data set to define or generate a trained machine learning model. The method may include obtaining one or more parameters for a pipeline operation that correspond to (a) operation of one or more pumps for each of the one or more sections of the pipeline, (b) the DRA to be utilizedfor the pipeline operation, and (c) a hydrocarbon to flow through each of the one or more sections of the pipeline during the pipeline operation. The method may include determining, based on application of the one or more parameters to the trained machine learning model, an amount of the DRA to inject into each of the one or more sections of the pipeline at one or more selected times. The method may include, in response to determination of the amount of the DRA, injecting the amount of the DRA into the selected section of the one or more sections of the pipeline. In an embodiment, if pump power is determined, then the pump power may be adjusted and / or pumps may be shut-down or shut-off.

[0014] In an embodiment, the training data set includes one or more of parameters of the hydrocarbon, a start time that the distinct hydrocarbon flow begins in for the selected section of the one or more sections of the pipeline, an end time that the distinct hydrocarbon flow begins in for the selected section of the one or more sections of the pipeline, a previous amount of the DRA injected into the selected section of the one or more sections of the pipeline, properties of the DRA injected into the selected section of the one or more sections of the pipeline, a pipeline diameter for the selected section of the one or more sections of pipeline, properties of the pump for the selected section of the one or more sections of the pipeline, a power amount used for the pump corresponding to the selected section of the one or more sections of the pipeline, a viscosity of the distinct hydrocarbon flow, a gravity of the distinct hydrocarbon flow, or a flow rate of the distinct hydrocarbon flow. A grade code defines the properties of the hydrocarbon. In another embodiment, the determining of the amount of the DRA to inject into each of the one or more sections of the pipeline at the one or more selected times is further based on application of an amount of drag reduction input for the pipeline operation.

[0015] In another embodiment, the method may include, prior to training the machine learning model, determining whether the training data set includes an amount of data that exceeds a first selected threshold; and, if the amount of data exceeds the first selected threshold, preprocessing the training data set. Preprocessing the training data set my include normalizing the training data set thereby defining a normalized training data set; and splitting the training data set into a training set and a test set. The method may include, subsequent to preprocessing the training data set, training the machine learning model with the training set; testing the machine learning model with the test set; determining if the trained machine learning model performance during testing meets a second threshold; and,if the performance does not meet the second threshold, re-training the machine learning model with a different portion of the training data set.

[0016] In an embodiment, the method may include determining, based on application of the one or more parameters to the trained machine learning model, an amount of drag reduction for the pipeline operation. Determination of one or more of (i) an amount of DRA to inject into each of the one or more sections of the pipeline at one or more selected times or (ii) an amount of drag reduction for the pipeline operation may further based on an input specifying an output of the trained machine learning model.

[0017] In an embodiment, the one or more parameters may be obtained from one or more sensors or analyzers positioned at each of the one or more sections of the pipeline.

[0018] Another embodiment of the disclosure is directed to a method to determine an amount of DRA to inject into one or more sections of pipeline. The method may include generating a training data set based a plurality of data sets that each correspond to a distinct hydrocarbon flow for a selected section of the one or more sections of the pipeline. The method may include training a machine learning model with the training data set that corresponds to the selected section of the one or more sections of the pipeline to define or generate a trained machine learning model for each section of the one or more sections of the pipeline. The method may include obtaining one or more parameters for a pipeline operation that correspond to (a) operation of one or more pumps for each of the one or more sections of the pipeline, (b) the DRA to be utilized for the pipeline operation, and (c) a hydrocarbon to flow through each of the one or more sections of the pipeline during the pipeline operation. The method may include determining, for each section of the one or more sections of the pipeline and based on application of the one or more parameters corresponding to a selected section of the one or more sections of the pipeline to the trained machine learning model for the selected section of the one or more sections of the pipeline, an amount of the DRA to inject into each of the one or more sections of the pipeline at one or more selected times. The method may include, in response to determination of the amount of the DRA, injecting the amount of the DRA into the selected section of the one or more sections of the pipeline.

[0019] Another embodiment of the disclosure is directed to a method to determine an amount of DRA to inject into one or more sections of pipeline. The method may include generating a training data set based on one or more data sets that each correspond to a distinct hydrocarbon flow for a selected section of the one or more sections of the pipeline. The method may include training a machine learning model with the training data set todefine or generate a trained machine learning model. The method may include receiving a plurality of parameters for a pipeline operation that correspond to (a) operation of one or more pumps for each of the one or more sections of the pipeline, (b) the DRA to be utilized for the pipeline operation, and (c) a hydrocarbon to flow through each of the one or more sections of the pipeline during the pipeline operation. The method may include determining, based on application of the one or more parameters to the trained machine learning model, an amount of DRA to inject into each of the one or more sections of the pipeline at one or more selected times. The method may include injecting the amount of DRA into the selected section of pipeline.

[0020] In an embodiment, the method may include, during the pipeline operation, determining updated one or more parameters; determining, based on application of the updated one or more parameters and one or more of the plurality of parameters to the trained machine learning model, an updated amount of DRA to inject into each of the one or more sections of the pipeline at one or more selected times; and injecting the updated amount of DRA into the selected section of pipeline. Determination of the updated amount of DRA may occur at selected time intervals and / or in real-time. In an embodiment, the plurality of parameters may be received from one or more sensors or analyzers associated with one or more of (a) each of the one or more pumps, (b) each of one or more DRA injection devices, and (c) each of the one or more sections of the pipeline.

[0021] Another embodiment of the disclosure is directed to a method to determine an amount of DRA to inject into a section of pipeline. The method may include determining parameters. The parameters may include one or more of (a) properties of hydrocarbon flowing through the section of pipeline, (b) a current amount and properties of DRA injected into section of the pipeline, and (c) one or more of a speed or power of a pump for the section of pipeline. The method may include determining one or more of (i) an adjusted amount of DRA to inject into the section of the pipeline or (ii) a drag reduction amount for the section of the pipeline based on application of the parameters to a trained machine learning model. The method may include injecting the adjusted amount of DRA into the section of the pipeline.

[0022] In an embodiment, the parameters may be received from one or more sensors or analyzers associated with one or more of (a) each of the one or more pumps, (b) each of one or more DRA injection devices, and (c) each of the one or more sections of the pipeline. The drag reduction amount may comprise an amount of power for the pump of the section of the pipeline and the adjusted amount of DRA for the section of the pipeline or, in anotherembodiment, drag reduction agent effectiveness. Determination of the adjusted amount of DRA to inject into the section of the pipeline may be further based on user input including entry of the drag reduction amount for the section of pipeline. Determination of the parameters and the one or more of (i) the adjusted amount of DRA or (ii) the drag reduction amount occurs at selected time intervals during a pipeline operation.

[0023] In another embodiment, determination of the parameters or a portion of the parameters occurs in real-time at each of selected time intervals during pipeline operation. The parameters may include a prior amount of DRA included in a hydrocarbon flow from a previous section of the pipeline to the section of the pipeline.

[0024] Another embodiment of the disclosure is directed to a method to determine an amount of drag reduction agent (DRA) to inject into a plurality of sections of pipeline. Each step in the method may be performed for each section of the plurality of sections of the pipeline. Such a method may include generating a training data set based on a plurality of data sets that each correspond to a distinct hydrocarbon flow for a selected section of the plurality of sections of the pipeline. The method may include training a machine learning model with the training data set to define or generate a trained machine learning model. The method may include obtaining one or more parameters for a pipeline operation that correspond to (a) operation of one or more pumps for the selected section of the plurality of sections of the pipeline, (b) the DRA to be utilized for the pipeline operation, and (c) a hydrocarbon to flow through the selected section of the plurality of sections of the pipeline during the pipeline operation. The method may include determining, based on application of the one or more parameters to the trained machine learning model for the selected section of the plurality of sections of the pipeline, an amount of the DRA to inject into the selected section of the plurality of sections of the pipeline at one or more selected times. The method may include in response to determination of the amount of the DRA, injecting the amount of the DRA into the selected section of the one or more sections of the pipeline.

[0025] Another embodiment of the disclosure is directed to a method to determine one or more of an amount of drag reduction agent (DRA) to inject into a section of pipeline or a drag reduction amount for the section of pipeline. Each step in the method may be performed for each section of the plurality of sections of the pipeline. The method may include obtaining one or more parameters for a pipeline operation that correspond to (a) operation of one or more pumps for a selected section of the plurality of sections of the pipeline, (b) the DRA to be utilized for the pipeline operation, or (c) a hydrocarbon to flow through the selected section of the plurality of sections of the pipeline during the pipelineoperation. The method may include determining, based on application of the one or more parameters to the trained machine learning model for the selected section of the plurality of sections of the pipeline, one of (a) an amount of the DRA to inject into the selected section of the plurality of sections of the pipeline at one or more selected times or (b) an amount of drag reduction for the selected section of the plurality of sections of the pipeline. The method may include one of (i) in response to determination of the amount of the DRA, injecting the amount of the DRA into the selected section of the plurality of sections of the pipeline; or (i) in response to determination of the amount of the drag reduction, injecting a second amount of the DRA into the selected section of the plurality of sections of the pipeline.

[0026] In an embodiment, the one or more parameters may include flow rate, grade code, viscosity, gravity, or temperature of fluid flowing through the selected section of the plurality of sections of the pipeline or length of the selected section of the plurality of sections of the pipeline. In another embodiment, the one or more parameters may include flow rate, gravity, or viscosity of fluid flowing through the selected section of the plurality of sections of the pipeline.

[0027] Another embodiment of the disclosure is directed to a system for determining an amount of DRA to inject into one or more sections of pipeline. The system may include one or more sensors positioned at one or more of on, within, or proximal to one or more components and configured to measure parameters associated with each of the one or more components including one or more of (a) each of the one or more sections of pipeline or (b) each of one or more pumps corresponding to one of the one or more sections of pipeline. The system may include a DRA modeling circuitry. The DRA modeling circuitry may be configured to obtain parameters measured by the one or more sensors. The DRA modeling circuitry may be configured to apply the parameters to a trained machine learning model. The DRA modeling circuitry may be configured to determine an adjusted amount of DRA to inject into each section of the pipeline based on application of the one or more parameters to the trained machine learning model. The DRA modeling circuitry may be configured to initiate injection of the adjusted amount of DRA into each section of the pipeline.

[0028] The system may include DRA injection devices in signal communication with the DRA modeling circuitry and each one of the DRA injection devices corresponding to one of each section of the pipeline. Initiation of injection of the adjusted amount of DRA for a section of the pipeline may include the DRA modeling circuitry communicating a signal to a corresponding DRA injection device. The signal may indicate the amount of DRA toinject and causing the corresponding DRA injection device to begin injection of the amount ofDRA.

[0029] The DRA modeling circuitry may be further configured to obtain predetermined parameters for each section of the pipeline. Determination of the adjusted amount ofDRA to inject into each section of the pipeline may further be based on application of the predetermined parameters to the trained machine learning model.

[0030] The DRA modeling circuitry may be further configured to obtain the parameters and determine the adjusted amount of DRA at preselected time intervals during a pipeline operation.

[0031] The system may include one or more pumps. Each pump may be positioned in fluid flow with each section of the pipeline and may be configured to cause a fluid to flow through a section of the pipeline at a selected flow rate. The fluid may comprise one or more of a hydrocarbon-based fluid or a renewable hydrocarbon-based fluid.

[0032] Another embodiment of the disclosure is directed to a system for determining an amount ofDRA to inject into a hydrocarbon flow of one or more sections of pipeline. The system may include a plurality of pipeline sections. The system may include a plurality of pumps in fluid communication. Each one of the plurality of pumps may correspond to one of the plurality of pipeline sections. The system may include a plurality ofDRA containers to store DRA. Each one of the plurality of DRA containers may correspond to one of the plurality of pipeline sections. The system may include a plurality ofDRA injection devices. Each one of the plurality of DRA injection devices may connect to one of the plurality of DRA containers and one of the plurality of pipeline sections. Each one of the plurality of DRA injection devices may be configured to inject an amount of the DRA into each one of the plurality of pipeline sections. The system may include one or more sensors positioned to measure parameters at one or more of on, within, or proximal to one or more of (a) the plurality of pipeline sections, (b) the plurality of pumps, (c) the plurality ofDRA containers, or (d) the plurality of DRA injection devices. The system may include a DRA modeling circuitry. The DRA modeling circuitry may be configured to obtain the parameters measured by the one or more sensors. The DRA modeling circuitry may be configured to apply the parameters to a trained machine learning model. The DRA modeling circuitry may be configured to determine an adjusted amount of DRA to inject into each section of the pipeline based on application of the parameters to the trained machine learning model. The DRA modeling circuitry may be configured to initiate injection of the adjusted amountof DRA into each section of the pipeline via a signal transmitted to one of the plurality of DRA injection devices and indicative of the adjusted amount of DRA.

[0033] The parameters may include one or more of (a) properties of hydrocarbon flowing through each of the plurality of pipeline sections, (b) a current amount and properties of DRA injected into each of the plurality of pipeline sections, (c) one or more of a speed or power of each of the plurality of pumps for each of the plurality of pipeline sections, or (d) an available amount of DRA in each of the plurality of DRA containers.

[0034] Another embodiment of the disclosure is directed to a controller for determining an amount of DRA to inject into one or more sections of pipeline. The controller may include a first plurality of inputs each in signal communication with one or more of (a) one or more sensors configured to measure a first set of parameters or (b) a user interface configured to receive a second set of parameters. The controller may be configured to apply one or more of the first set of parameters or the second set of parameters to a trained machine learning model. The controller may be configured to determine an adjusted amount of DRA to inject into a selected section of pipeline based on application of the one or more of the first set of parameters or the second set of parameters to the trained machine learning model. The controller may include a first output in signal communication with a DRA injector device. The controller may be configured to cause the DRA injector device to inject the adjusted amount of DRA into the selected section of pipeline.

[0035] In an embodiment, the controller may include a second output in signal communication with a pump configured to cause hydrocarbon to flow through the selected section of pipeline at a selected flowrate. The controller may be configured to determine an adjusted flowrate of the hydrocarbon. The controller may be configured to cause the pump, if the pump is a VFD or VSD pump, to adjust the selected flowrate to the adjusted flowrate of the hydrocarbon. In another embodiment, the controller may be configured to cause a pump to shut-down, shut-off, or cease operation. Such adjustment or shut-down may be based on application of the first set of parameters and / or second set of parameters to the trained machine learning model.

[0036] Still other aspects and advantages of these embodiments and other embodiments, are discussed in detail herein. Moreover, it is to be understood that both the foregoing information and the following detailed description provide merely illustrative examples of various aspects and embodiments, and are intended to provide an overview or framework for understanding the nature and character of the claimed aspects and embodiments. Accordingly, these and other objects, along with advantages and features of the presentdisclosure herein disclosed, will become apparent through reference to the following description and the accompanying drawings. Furthermore, it is to be understood that the features of the various embodiments described herein are not mutually exclusive and may exist in various combinations and permutations.BRIEF DESCRIPTION OF THE DRAWINGS

[0037] These and other features, aspects, and advantages of the disclosure will become better understood with regard to the following descriptions, claims, and accompanying drawings. It is to be noted, however, that the drawings illustrate only several embodiments of the disclosure and, therefore, are not to be considered limiting of the scope of the disclosure.

[0038] FIG. l is a simplified diagram of a drag reduction agent modeling system, according to an embodiment of the disclosure.

[0039] FIG. 2 is a simplified diagram that illustrates an apparatus for determining an amount of drag reduction agent to inject into a pipeline and / or an amount or percentage of drag reduction, according to an embodiment of the disclosure.

[0040] FIG. 3 is a schematic diagram of a drag reduction agent modeling and adjustment system, according to an embodiment of the disclosure.

[0041] FIG. 4 is a simplified diagram that illustrates a control system for determining an amount of drag reduction agent to inject into a pipeline and / or an amount or percentage of drag reduction, according to an embodiment of the disclosure.

[0042] FIG. 5A and FIG. 5B illustrate a block diagram of graphical user interfaces to determine an amount of drag reduction agent to inject into a pipeline, according to an embodiment of the disclosure.

[0043] FIG. 6 is a flow diagram for training a machine learning model, according to an embodiment of the disclosure.

[0044] FIG. 7 is a flow diagram for using a trained machine learning model to determine an amount of drag reduction agent and / or an amount or percentage of drag reduction, according to an embodiment of the disclosure.DETAILED DESCRIPTION

[0045] So that the manner in which the features and advantages of the embodiments of the systems and methods disclosed herein, as well as others that will become apparent, may be understood in more detail, a more particular description of embodiments of systems andmethods briefly summarized above may be had by reference to the following detailed description of embodiments thereof, in which one or more are further illustrated in the appended drawings, which form a part of this specification. It is to be noted, however, that the drawings illustrate only various embodiments of the systems and methods disclosed herein and are therefore not to be considered limiting of the scope of the systems and methods disclosed herein as it may include other effective embodiments as well.

[0046] The present disclosure provides embodiments of systems and methods for adjusting an amount of drag reduction agent (DRA) injected into hydrocarbon liquids flowing through a pipeline. “Hydrocarbon liquids” as used herein, may refer to petroleum liquids, renewable liquids, and other hydrocarbon based liquids. “Petroleum liquids” as used herein, may refer to liquid products containing crude oil, petroleum products, and / or distillates or refinery intermediates. For example, crude oil contains a combination of hydrocarbons having different boiling points that exists as a viscous liquid in underground geological formations and at the surface. Petroleum products, for example, may be produced by processing crude oil and other liquids at petroleum refineries, by extracting liquid hydrocarbons at natural gas processing plants, and by producing finished petroleum products at industrial facilities. Refinery intermediates, for example, may refer to any refinery hydrocarbon that is not crude oil or a finished petroleum product (e.g., such as gasoline), including all refinery output from distillation (e.g., distillates or distillation fractions) or from other conversion units. In some non-limiting embodiments of systems and methods, petroleum liquids may include heavy blend crude oil used at a pipeline origination station. Heavy blend crude oil is typically characterized as having an American Petroleum Institute (API) gravity of about 30 degrees or below. In other embodiments, the petroleum liquids may include lighter blend crude oils, for example, having an API gravity of greater than 30 degrees. “Renewable liquids” as used herein, may refer to liquid products containing plant and / or animal derived feedstock. Further, the renewable liquids may be hydrocarbon based. For example, a renewable liquid may be a pyrolysis oil, oleaginous feedstock, biomass derived feedstock, or other liquids, as will be understood by those skilled in the art. The API gravity of renewable liquids may vary depending on the type of renewable liquid.

[0047] In view of the foregoing, Applicant has recognized these problems and others in the art, and has recognized a need for enhanced systems and methods for determining an amount of DRA to inject at one or more sections of pipeline. Particularly, the present disclosure relates to systems and methods for determining an amount of DRA to inject intoone or more sections of pipeline used for transporting hydrocarbon liquids and / or renewable liquids and, particularly, for determining and adjusting one or more of an amount of DRA or a drag reduction amount or percentage for a selected section of pipeline using one or more trained machine learning models.

[0048] The disclosure herein provides embodiments of systems and methods for determining and adjusting an amount of DRA to inject into each section of pipeline. In an embodiment, the systems and methods may include a computing device, apparatus and / or controller to obtain various data points and / or parameters to train a machine learning model and to apply to a trained machine learning model. The system may include, for example, for each section of a pipeline, a pump, a DRA injection device, a DRA tank or container, and / or a plurality of sensors and / or analyzers. In an example, each section of the pipeline may include a starting location, defined by a pump or pumping station, and an end location, defined by the beginning of a subsequent pump or pumping station. Each section of the pipeline may be different in or similar in length than preceding and / or subsequent sections of the pipeline. For example, a section of pipeline may comprise a length of about 10 miles, about 20 miles, about 40 miles, about 60 miles, about 10 miles to about 60 miles, and / or other lengths as will be understood by one skilled in the art. The DRA injection device and DRA tank or container may be positioned proximate the pump or pumping station. In other embodiments, each section of the pipeline may include additional DRA injection devices and DRA tanks or containers. In another embodiment, the DRA injection device may comprise a valve, a pump, or some combination thereof. The DRA injection device may be in fluid communication with the DRA tank or container at a first end and with a section of the pipeline at a second end, thus allowing DRA to be injected from the DRA tank or container into the section of the pipeline. In embodiments, each DRA tank or container may include a type of DRA and / or a DRA with selected properties. The DRA may include a selected A and B coefficient, indicating how the DRA may interact with other liquids. In other embodiments, each DRA tank or container may include different types of DRA. In other words, one, two, or more DRA tanks may include a first type of DRA, while another one, two or more DRA tanks may include a second type of DRA.

[0049] As a pipeline operation begins, hydrocarbon may begin flowing from a source to a first pump or pump station. As the hydrocarbon flows through the pump or pump stations, the pump may raise the hydrocarbon liquids to high pressure for enabling travel of the liquid at long distances. Prior to such a pipeline operation, the systems and methods may include applying one or more parameters to a trained machine learning model. Such anapplication to the trained machine learning models may define, determine, or generate a value indicative of how much DRA should be injected into a section of the pipeline, an amount of drag reduction for the pipeline operation at the section of the pipeline, and / or what the power and / or speed (including, in an embodiment, whether the pump should cease operating or begin operating) of a pump should be set to for the section of the pipeline. Such outputs may be determined based on the parameters applied to the trained machine learning model and / or based on an input indicative of the type of output desired. Upon output of such values, the systems and methods may adjust one or more of the amount of DRA to be injected or being injected into the section of the pipeline and / or, in an some embodiments, a speed and / or power of the pump for the section of the pipeline.

[0050] The embodiments described above may be iterative for each section of pipeline. For example, at the beginning of or prior to initiation of a pipeline operation, the obtained parameters may be applied to the trained machine learning model for each section of the pipeline to determine an initial amount of DRA to inject into each section of the pipeline and / or an amount of drag reduction for the pipeline operation at the section of the pipeline. In other embodiments, the determination may include a power to set the pump to for each section of the pipeline and / or the operating state of the pump for each section of the pipeline. The operating state of a pump may include whether the pump is actively operating or shut-down or shut-off. As the pipeline operation continues, parameters measured or obtained during the pipeline operation may be utilized to adjust the DRA amount (and, in some embodiments, pump power, speed, and / or operating state) in real-time (for example, based on application of the parameters to the trained machine learning model). Thus, adjustments may occur in real-time over the course of a pipeline operation.

[0051] In yet other embodiments, the system may include a plurality of controllers. Each one of the plurality of controllers may be positioned at each one of the sections of pipeline. Further, each controller may include the trained machine learning model. The controller may store selected parameters specific to a corresponding section of the pipeline. In other embodiments, the controller may store a selected trained machine learning model that has been trained using parameters specific to that corresponding section of pipeline. In yet another embodiment, a supervisory controller or computing device may receive parameters for each section and apply those parameters to the trained machine learning model.

[0052] In an embodiment, the parameters noted above may include various measurements, properties, and / or indicators obtained in real-time and / or that are static or change less frequent rates. For example, the parameters may include pipeline diameter for each sectionof the pipeline, the direction that the pipeline runs (for example, at a downward angle, upward angle, or horizontally), the location of the pipeline (for example, underground or aboveground), any known internal anomalies of the pipeline (for example, rough sections of the pipeline), properties of hydrocarbon for a pipeline operation (for example, a grade or grade code of the hydrocarbon and / or a specification indicating various parameters of the hydrocarbon), a viscosity of the hydrocarbon, a gravity of the hydrocarbon, an amount of the hydrocarbon, a flow rate of the hydrocarbon, a power or speed setting for each pump, an amount of DRA to be injected into a section of the pipeline, the properties of DRA to be utilized, temperature within various locations of each section of the pipeline, and / or the pressure drop from the beginning to the end of each section of the pipeline, among other parameters. In an embodiment, to obtain an amount of DRA to inject at a section, the system may obtain a desired drag reduction amount and hydrocarbon grade code. In other embodiments, additional factors or parameters may be utilized. In yet another embodiment, other parameters may be utilized to obtain a desired drag reduction, for example an amount of DRA set to be injected in pipeline.

[0053] As noted, the systems and methods may utilize a trained machine learning model. The systems and methods may train the trained machine learning model. To begin the training process, the systems and methods may first determine or generate a training data set. The systems and methods may include separating the historical data set into distinct data sets based on a distinct hydrocarbon flowing through the pipeline at a selected time. Each of the distinct data sets may include one or more of the parameters described above. Once a selected threshold amount of distinct data sets have been determined, the distinct data sets may be split into a training set and a testing set. The system may train the machine learning model using the training set and then test the trained machine learning model with the testing set. If the output of the trained machine learning model falls below a second selected threshold, the system may retrain the trained machine learning model with a different portion of the distinct data sets and / or with additional historical data. Once the trained machine learning model meets the second selected threshold, the system may utilize the trained machine learning model during pipeline operations for each section of the pipeline. In an embodiment, the trained machine learning model, as noted, may correspond to a section of the pipeline. In another embodiment, the trained machine learning model may not be pipeline section specific. In other words, the trained machine learning model may be utilized for any section of the pipeline and / or the full length of the pipeline.

[0054] By utilizing the trained machine learning model, the systems and methods described herein may determine a specific amount of DRA to be utilized in each section of pipeline and / or an amount of drag reduction for each section of the pipeline during a pipeline operation. This determination may occur prior to initialization of the pipeline operation and during the pipeline operation. Thus, as various parameters change over time (for example, hydrocarbon viscosity and / or DRA build up from previous sections of pipeline, among other factors), the systems and methods may continue to adjust DRA amount to ensure balance between injected DRA and pump power, as well as that a selected drag reduction is maintained.

[0055] Thus, rather than utilizing a random amount of DRA for any section of the pipeline, the systems and methods described herein enable precise amounts of DRA (for example, at parts per million (PPM) in relation to the amount of hydrocarbon flowing through the pipeline) to be utilized from initialization to the end of the pipeline operation. Further, selected amounts of drag reduction may be maintained, as drag reduction changes over time due to various factors or parameters. Additionally, the trained machine learning model may be trained such that minimal input to the trained machine learning model may determine the precise amount of DRA for selected grades or grade codes of hydrocarbons.

[0056] FIG. 1 illustrates a simplified diagram of a drag reduction agent modeling system, according to an embodiment of the disclosure. As illustrated, the DRA modeling system 102 may connect to various components. In embodiments, the DRA modeling system 102 may connect to components or devices associated with and / or corresponding to each of a plurality of pipeline sections. In such embodiments, the DRA modeling system 102 may connect to a pump (for example, pump 116A, 116B, and up to 116N), a DRA injection device (for example, DRA injection device 114A, 114B, and up to 114N), one or more sensors and / or analyzers 112A, 112B, and up to 112N, and / or a computing device 118. The DRA modeling system 102 may further connect to other components or devices. For example, the DRA modeling system may connect to a DRA tank or container, one or more controllers positioned at a pump station, and / or other devices. In such embodiments, the DRA modeling system 102 may include a communications circuitry or module. The communications circuitry or module may enable the DRA modeling system 102 to communicate with the other components and devices described herein. For example, the DRA modeling system 102 may send signals indicative of an amount of DRA to inject into a pipeline to a DRA injection device 114 A, 114B, and up to 114N.

[0057] While not illustrated in FIG. 1, a section of the pipeline or pipeline section may include one or more pumps 116A, 116B, and up to 116N, one or more DRA injection device 114A, 114B, and up to 114N, and / or one or more sensors and / or analyzers 112A, 112B, and up to 112N. Each section of the pipeline may be in fluid communication with a preceding and / or subsequent section of the pipeline (as illustrated in FIG. 3). Each pipeline section may include various parameters associated with that pipeline section, such as pipeline diameter, pipeline section length, pump characteristics and / or power, DRA injection device characteristics, other pipe parameters, and / or other geographic and / or environment characteristics (for example, a portion of or all of a section of pipeline may be underground, underwater, above ground, running at a downward angle, running at an upward angle, and / or running substantially horizontally, among other characteristics). Such factors may affect the power utilized by a pump in that specific section, as well as the amount of DRA used for that section. Further, each section of the pipeline may include a controller and / or computing device. In such embodiments, each controller and / or computing device may include the functionality of the DRA modeling system 102. In another embodiment, the controller and / or computing device at each section of the pipeline may be in signal communication with the DRA modeling system 102.

[0058] The DRA modeling system 102 may include a processor 104 and a memory 106 or non-transitory machine-readable storage medium storing instructions executable by the processor 104. In some examples, the DRA modeling system 102 may be a computing device. The term “computing device” is used herein to refer to any one or all of programmable logic controllers (PLCs), programmable automation controllers (PACs), industrial computers, servers, virtual computing device or environment, desktop computers, personal data assistants (PDAs), laptop computers, tablet computers, smart books, palm-top computers, personal computers, smartphones, virtual computing devices, cloud based computing devices, and similar electronic devices equipped with at least a processor and any other physical components necessarily to perform the various operations described herein. Devices such as smartphones, laptop computers, and tablet computers are generally collectively referred to as mobile devices.

[0059] The term “server” or “server device” is used to refer to any computing device capable of functioning as a server, such as a master exchange server, web server, mail server, document server, or any other type of server. A server may be a dedicated computing device or a server module (e.g., an application) hosted by a computing device that causes the computing device to operate as a server. A server module (e.g., server application) maybe a full function server module, or a light or secondary server module (e.g., light or secondary server application) that is configured to provide synchronization services among the dynamic databases on computing devices. A light server or secondary server may be a slimmed-down version of server type functionality that can be implemented on a computing device, such as a smart phone, thereby enabling it to function as an Internet server (e.g., an enterprise e-mail server) only to the extent necessary to provide the functionality described herein.

[0060] As used herein, a “non-transitory machine-readable storage medium” or “memory” may be any electronic, magnetic, optical, or other physical storage apparatus to contain or store information such as executable instructions, data, and the like. For example, any machine-readable storage medium described herein may be any of random access memory (RAM), volatile memory, non-volatile memory, flash memory, a storage drive (e.g., hard drive), a solid state drive, any type of storage disc, and the like, or a combination thereof. The memory may store or include instructions executable by the processor.

[0061] As used herein, a “processor” or “processing circuitry” may include, for example one processor or multiple processors included in a single device or distributed across multiple computing devices. The processor (such as, processor 104 shown in FIG. 1 or processing circuitry 202 shown in FIG. 2) may be at least one of a central processing unit (CPU), a semiconductor-based microprocessor, a graphics processing unit (GPU), a field- programmable gate array (FPGA) to retrieve and execute instructions, a real time processor (RTP), other electronic circuitry suitable for the retrieval and execution instructions stored on a machine-readable storage medium, or a combination thereof.

[0062] As illustrated, the DRA modeling system 102 may include one or more different instructions. For example, the DRA modeling system 102 may include DRA modeling instructions 107. The DRA modeling instructions 107, when executed by the processor 104, may cause the DRA modeling system 102 to obtain various parameters for one or more sections of the pipeline. Once the DRA modeling system 102 has obtained the parameters, the DRA modeling instructions 107 may apply the parameters to a trained machine learning model stored within the memory 106 or stored in remote memory or storage accessible by the DRA modeling system 102. Applying the parameters to the trained machine learning model may determine or generate a value or values. The value or values may be indicative of how much DRA should be injected into a section of the pipeline and / or an amount of drag reduction for the pipeline operation at the section of the pipeline. In an embodiment, the amount of drag reduction for the pipeline operation may include an amount of powerfor the pump of the section of the pipeline and the adjusted amount of DRA for the section of the pipeline or, in another embodiment, drag reduction agent effectiveness for a selected section of the pipeline. In embodiments, the value or values may also be indicative of what the power, speed, and / or operating state of a pump should be set to.

[0063] Once the values have been determined, the processor 104 may execute the DRA adjustment instructions 108, based upon the output of the trained machine learning model. For example, if the trained machine learning model determines a value indicative of an amount of DRA to inject into the section of pipeline, then the DRA adjustment instructions 108 may be executed.

[0064] If the DRA adjustment instructions 108 are executed, then the DRA modeling system 102 may transmit a signal to a corresponding DRA injection device (for example, DRA injection device 114A , 114B, and up to 114N). The signal may be indicative of an adjustment to reach the adjusted amount of DRA to be inj ected into a section of the pipeline. The signal may cause the DRA injection device (for example, DRA injection device 114A , 114B, and up to 114N) to begin injecting the indicated amount of DRA. If pump speed, power, and / or operating state are included in the DRA adjustment instructions 108 or other instructions stored in memory 106, when those instructions are executed, then the DRA modeling system 102 may transmit a signal to a corresponding pump (for example, pump 116A , 116B, and up to 116N). The signal may be indicative of an adjustment to reach the adjusted pump speed and / or power or for an operating state that a pump is to be set to for a corresponding section of the pipeline. The signal may cause the pump to adjust to the indicated speed and / or power or cause the pump to transition to the indicated operating state.

[0065] In an embodiment, the pump 116A, 116B, and up to 116 N may be a set speed / frequency pump or a variable speed / variable frequency drive (VFD) pump. If the pump is a VFD pump, then the pump may speed up or slow down to increase or decrease, respectively, pressure and / or flow rate. In an embodiment, the DRA injection device 114A, 114B, and up to 114N may include a pump, a valve, and / or other flow control or injection device. The DRA injection device 114A, 114B, and up to 114N may be configured to control a flow rate of DRA into the section of pipeline at parts per million (PPM) in relation to the hydrocarbon liquid and / or renewable liquid flowing through the pipeline.

[0066] FIG. 2 is a simplified diagram that illustrates an apparatus 200 for determining an amount of DRA to inject into a pipeline and / or to determine an amount drag reduction for a section of the pipeline, according to an embodiment of the disclosure. FIG. 2 is asimplified diagram that illustrates an apparatus 200 for determining an amount of DRA to inject into a pipeline and / or an amount of power for a pump to utilize, according to an embodiment of the disclosure. Such an apparatus 200 may be comprised of a processing circuitry 202, a memory 204, a communications circuitry 206, a DRA modeling circuitry 208, and a DRA injection circuitry 210, each of which will be described in greater detail below. While the various components are illustrated in FIG. 2 as being connected with processing circuitry 202, it will be understood that the apparatus 200 may further comprise a bus (not expressly shown in FIG. 2) for passing information amongst any combination of the various components of the apparatus 200. The apparatus 200 may be configured to execute various operations described herein, such as those described above in connection with FIG. 1 and below in connection with FIGS. 3-6.

[0067] The processing circuitry 202 (and / or co-processor or any other processor assisting or otherwise associated with the processor) may be in communication with the memory 204 via a bus for passing information amongst components of the apparatus. The processing circuitry 202 may be embodied in a number of unusual ways and may, for example, include one or more processing devices configured to perform independently. Furthermore, the processor may include one or more processors configured in tandem via a bus to enable independent execution of software instructions, pipelining, and / or multithreading.

[0068] The processing circuitry 202 may be configured to execute software instructions stored in the memory 204 or otherwise accessible to the processing circuitry 202 (e.g., software instructions stored on a separate storage device). In some cases, the processing circuitry 202 may be configured to execute hard-coded functionality. As such, whether configured by hardware or software methods, or by a combination of hardware with software, the processing circuitry 202 represents an entity (for example, physically embodied in circuitry) capable of performing operations according to various embodiments of the present disclosure while configured accordingly. Alternatively, as another example, when the processing circuitry 202 is embodied as an executor of software instructions, the software instructions may specifically configure the processing circuitry 202 to perform the algorithms and / or operations described herein when the software instructions are executed.

[0069] Memory 204 is non-transitory and may include, for example, one or more volatile and / or non-volatile memories. In other words, for example, the memory 204 may be an electronic storage device (for example, a computer readable storage medium). The memory 204 may be configured to store information, data, content, applications, softwareinstructions, or the like, for enabling the apparatus 200 to carry out various functions in accordance with example embodiments contemplated herein.

[0070] The communications circuitry 206 may be any means such as a device or circuitry embodied in either hardware or a combination of hardware and software that is configured to receive and / or transmit data from / to a network and / or any other device, circuitry, or module in communication with the apparatus 200. In this regard, the communications circuitry 206 may include, for example, a network interface for enabling communications with a wired or wireless communication network. For example, the communications circuitry 206 may include one or more network interface cards, antennas, buses, switches, routers, modems, and supporting hardware and / or software, or any other device suitable for enabling communications via a network. Furthermore, the communications circuitry 206 may include the processing circuitry for causing transmission of such signals to a network or for handling receipt of signals received from a network. The communications circuitry 206, in an embodiment, may enable reception of parameters from various components, devices, and / or sensors (for example, pumps, DRA injection devices, analyzers, sensors, and / or other components), as well as communication of instructions and / or signals indicative of adjustment to those components and / or devices.

[0071] The apparatus 200 may include a DRA modeling circuitry 208 configured to obtain parameters from one or more components, devices, sensors, and / or analyzers and / or apply those parameters to a trained machine learning model. Obtaining the parameters from the one or more components, devices, sensors, and / or analyzers may occur periodically, at selected times, continuously, or substantially continuously. The DRA modeling circuitry 208 may poll the components, devices, sensors, and / or analyzers to obtain such parameters or, in an embodiment, receive the parameters without polling. The DRA modeling circuitry 208 may obtain the parameters via the communications circuitry 206. Application of the parameters to the trained machine learning model may determine an output. The output may be indicative of an adjustment to the amount of DRA to be injected into the pipeline and / or an amount of drag reduction for a section of pipeline. In another embodiment, the DRA modeling circuitry 208 may train the trained machine learning model prior to use. In such embodiments, the DRA modeling circuitry 208 may obtain historical data, preprocess the historical data, and then train and test the machine learning model. In yet another embodiment, after a pipeline operation (in other words, transport of a hydrocarbon liquid and / or renewable liquid), the DRA modeling circuitry 208 may re-train or refine the trained machine learning model, based on the results of the pipeline operation. The DRA modelingcircuitry 208 may utilize processing circuitry 202, memory 204, or any other hardware component included in the apparatus 200 to perform these operations, as described in connection with FIGS. 3-6 below. The output of the DRA modeling circuitry 208 may be transmitted to other circuitry of the apparatus 200 (such as the DRA injection circuitry 210). In an embodiment, the DRA modeling circuitry 208 may also determine an adjustment to the power and / or speed that a pump operates at during the pipeline operation and / or an operating state that the pump is set to during the pipeline operation.

[0072] In addition, the apparatus 200 further comprises the DRA injection circuitry 210 that may cause one or more DRA injection devices to adjust the amount of DRA injected into a section of pipeline. The DRA injection circuitry 210 may utilize processing circuitry 202, memory 204, or any other hardware component included in the apparatus 200 to perform these operations, as described in connection with FIGS. 3-6 below. The DRA injection circuitry 210 may further utilize communications circuitry 206 to transmit the signal to the corresponding DRA injection device to cause the DRA injection device to inject an adjusted amount of DRA into the section of the pipeline. In an embodiment, the DRA injection circuitry 210 may cause one or more pumps corresponding to one or more sections of pipeline to adjust operating speed and / or power or transition to the indicated operating state (in other words, continue operation or cease operation).

[0073] Although components 202-212 are described in part using functional language, it will be understood that the particular implementations necessarily include the use of particular hardware. It should also be understood that certain of these components 202-212 may include similar or common hardware. For example, the DRA modeling circuitry 208, and the DRA injection circuitry 210 may, in some embodiments, each at times utilize the processing circuitry 202, memory 204, or communications circuitry 206, such that duplicate hardware is not required to facilitate operation of these physical elements of the apparatus 200 (although dedicated hardware elements may be used for any of these components in some embodiments, such as those in which enhanced parallelism may be desired). Use of the terms “circuitry,” with respect to elements of the apparatus therefore shall be interpreted as necessarily including the particular hardware configured to perform the functions associated with the particular element being described. Of course, while the terms “circuitry” should be understood broadly to include hardware, in some embodiments, the terms “circuitry” may in addition refer to software instructions that configure the hardware components of the apparatus 200 to perform the various functions described herein.

[0074] Although the DRA modeling circuitry 208, and the DRA injection circuitry 210 may utilize processing circuitry 202, memory 204, or communications circuitry 206 as described above, it will be understood that any of these elements of apparatus 200 may include one or more dedicated processors, specially configured field programmable gate arrays (FPGA), or application specific interface circuits (ASIC) to perform its corresponding functions, and may accordingly utilize processing circuitry 202 executing software stored in a memory or memory 204, communications circuitry 206 for enabling any functions not performed by special-purpose hardware elements. In all embodiments, however, it will be understood that the DRA modeling circuitry 208, and the DRA injection circuitry 210 are implemented via particular machinery designed for performing the functions described herein in connection with such elements of apparatus 200.

[0075] In some embodiments, various components of the apparatus 200 may be hosted remotely (e.g., by one or more cloud servers) and thus need not physically reside on the corresponding apparatus 200. Thus, some or all of the functionality described herein may be provided by third party circuitry. For example, a given apparatus 200 may access one or more third party circuitries via any sort of networked connection that facilitates transmission of data and electronic information between the apparatus 200 and the third party circuitries. In turn, that apparatus 200 may be in remote communication with one or more of the other components describe above as comprising the apparatus 200.

[0076] As will be appreciated based on this disclosure, example embodiments contemplated herein may be implemented by an apparatus 200 (or by a controller 402). Furthermore, some example embodiments (such as the embodiments described for FIGS. 1 and 3-6) may be a computer program product comprising software instructions stored on at least one non-transitory computer-readable storage medium (such as memory 204). Any suitable non-transitory computer-readable storage medium may be utilized in such embodiments, some examples of which are non-transitory hard disks, CD-ROMs, flash memory, optical storage devices, and magnetic storage devices. It should be appreciated, with respect to certain devices embodied by apparatus 200 as described in FIG. 2, that loading the software instructions onto a computing device or apparatus produces a specialpurpose machine comprising the means for implementing various functions described herein.

[0077] FIG. 3 is a schematic diagram of a DRA modeling and adjustment system, according to an embodiment of the disclosure. The system 300 may include various components and devices to control flow of various fluids from a source 303 to an end-userlocation. As illustrated, the system 300 may include a plurality of sections of pipeline 316A, 316B, and up to 316N. Each of the plurality of sections of pipeline 316A, 316B, and up to 316N may include a first end defined by a pump or pumping station and a second end defined by the input of a subsequent pump or pumping station. Each section of the pipeline 316A, 316B, and up to 316N may include a pump 304A, 304B, and up to 304N, a DRA injection device 312A, 312B, and up to 312N, a DRA tank 318A, 318B, and up to 318N, and / or various other components, such as sensors and / or analyzers. For example, a sensor and / or analyzer 302 may be positioned proximate the source 303 to measure one or more parameters of the hydrocarbon liquid and / or renewable liquid stored therein, such as grade code, viscosity, gravity, flow rate, temperature, density, and / or pressure. Further, each section may include a sensor and / or analyzer 308A, 308B, and up to 308N to measure one or more parameters (for example, grade code, flow rate, viscosity, gravity, temperature, and / or, in some embodiments, an amount of DRA remaining in the liquid) of the liquid flowing from the proximate pump 304A, 304B, and up to 304N. As noted, the sensor and / or analyzer 308A, 308B, and up to 308N may measure the amount of DRA remaining in a liquid. In such embodiments, as DRA passes through a pump, a portion, a substantial portion, or substantially all of the DRA may be destroyed. Some DRA may remain and, as such, that factor may be utilized when modeling an amount of DRA to inject into a section of the pipeline.

[0078] The system 300 may include another sensor and / or analyzer 310A, 310B, and up to 310N to measure one or more parameters, such as grade code, viscosity, flow rate, temperature, and / or pressure, among other parameters. Additionally, the system 300 may include a DRA sensor and / or analyzer 314A, 314B, and up to 314N to measure a flow rate of DRA.

[0079] In an embodiment, the DRA tank 318A, 318b, and up to 318N may include a tank and / or a container and may be positioned at a surface area proximate a corresponding pump and / or positioned on a skid or vehicle.

[0080] As illustrated in FIG. 3, the system may include a controller 301. In another embodiment, the system 300 may include a plurality of controllers, each controller corresponding to one of the sections of pipeline 316A, 316B, and up to 316N. Each of the plurality of controllers may include a trained machine learning model corresponding to a section of pipeline 316A, 316B, and up to 316N. In another embodiment, each of the plurality of controllers may facilitate communication between the various components and devices described herein and the controller 301. In another embodiment, the controller 301may directly communicate with each of the various components and devices described herein.

[0081] In an embodiment, prior to initiation of a pipeline operation, the controller 301 may receive parameters for the pipeline operation. The controller 301 may apply those parameters to a trained machine learning model to determine or generate an indicator. The indicator may indicate an amount of DRA to inject into each section of the pipeline 316A, 316B, and up to 316N. The parameters may include a grade and / or properties of hydrocarbon liquid and / or renewable liquid. Further, the parameters may include a desired drag reduction. In another embodiment, the indicator, in addition to or rather than indicating the amount of DRA, may indicate a power, speed, and / or operating state for each pump 304A, 304B, and up to 304N corresponding to each section of the pipeline 316A, 316B, and up to 316N. In yet another embodiment, the indicator may indicate a desired amount of drag reduction for the corresponding section of the pipeline 316A, 316B, and up to 316N.

[0082] In an embodiment, the type of indicator output by the trained machine learning model may be based on the parameters applied to the trained machine learning model. For example, an amount of DRA, in addition to other parameters, may be applied to the trained machine learning model to determine an indicator that indicates a desired amount of drag reduction. In yet another embodiment, an amount of drag reduction, in addition to other parameters, may be applied to the trained machine learning model to determine an indicator that indicates an amount of DRA to inject.

[0083] After initiation of the pipeline operation, the controller 301 may obtain one or more parameters from each of the sensors and / or analyzers positioned throughout the system 300. The controller 301 those parameters to the trained machine learning model to determine or generate an updated indicator. Each subsequent application of new parameters to the trained machine learning model may occur at a selected time interval and / or when a change in parameters is detected by the controller 301 via the measurements from the sensors and / or analyzers positioned throughout the system 300.

[0084] FIG. 4 is a simplified diagram that illustrates a control system 400 for determining an amount of DRA to inject into a pipeline and / or, in other embodiments, determine an amount of drag reduction for a section of the pipeline, according to an embodiment of the disclosure. As noted, control system 400 may include a controller 402. Further, controller 402 may connect to one or more pipe sensors and / or analyzers 418, one or more DRA sensors 420, one or more pumps 422, one or more DRA injection devices 424, and / or a user interface 416. The controller 402 may include memory 406 and one or more processors404. The memory 406 may store instructions executable by one or more processors 404. In an example, the memory 406 may be a non-transitory machine-readable storage medium. As noted, the memory 406 may store or include instructions executable by the processor 404.

[0085] As used herein, “signal communication” refers to electric communication such as hardwiring two components together or wireless communication, as understood by those skilled in the art. For example, wireless communication may be Wi-Fi®, Bluetooth®, ZigBee, or other near-field communications. In addition, signal communication may include one or more intermediate controllers or relays disposed between elements in signal communication.

[0086] The memory 406 may include or store model training instructions 408. When executed the model training instructions 408 may train a machine learning model. The model training instructions 408, to train the model, may first obtain a historical data set. The historical data set may include data from previous pipeline operations. Such data may include properties of the hydrocarbon liquid and / or renewable liquid, properties of the DRA utilized, an amount of DRA utilized, one or more pipeline characteristics, and / or one or more pump characteristics, among other data. The model training instructions 408, when executed by the processor 404, may preprocess the historical data set. Such preprocessing includes separating each pipeline operation based on a distinct hydrocarbon liquid and / or renewable liquid for a pipeline operation. In other words, may create a data set for each instance of a distinct hydrocarbon liquid and / or renewable liquid fully within a section of the pipeline. Once each pipeline operation is separated based on the hydrocarbon liquid and / or renewable liquid within the section of the pipeline, the model training instructions 408 may normalize the data. In other words, anomalies and / or outliers may be removed from the data set. In another embodiment, the model training instructions 408 may determine whether a selected threshold on an amount of data has been met, prior to further preprocessing. Once the data has been normalized, the data may be split into a training set and a test set. The model training instructions 408, when executed by the processor 404, may then train the machine learning model with the training set and then test the resulting trained machine learning model with the training set. Testing the machine learning model may determine or generate a performance indicator. If the performance indicator exceeds a selected threshold, then the trained machine learning model may be utilized during pipeline operations. If the performance indicator does not exceed the selected threshold, then the model training instructions 408 may obtain additional data and / or utilize different data inthe training set and testing set to retrain the machine learning model. Once a trained machine learning model is defined or determined, the trained machine learning model may be stored in memory 406.

[0087] The controller 402 may include DRA modeling instructions 410. The DRA modeling instructions 410 may, when executed by the processor 404, apply one or more parameters received from the one or more pipe sensors and / or analyzers 418, one or more DRA sensors 420, one or more pumps 422, one or more DRA injection devices 424, and / or a user interface 416 to the trained machine learning model. Such application of those parameters to the trained machine learning model may determine an adjusted amount of DRA for injection into one or more sections of pipeline. In another embodiment, the DRA modeling instructions 410 may determine an amount of drag reduction for the one or more sections of pipeline. Utilizing the output of the trained machine learning model, the DRA adjustment instructions 412 may cause a corresponding DRA injected device to inject the specified amount of DRA into the corresponding section of pipeline.

[0088] FIG. 5A and FIG. 5B illustrate a block diagram of graphical user interfaces to determine an amount of drag reduction agent to inject into a pipeline, according to an embodiment of the disclosure. Turning to FIG. 5A, a graphical user interface (GUI) 500 is provided that illustrates an embodiment of the DRA modeling system 102. As noted, the parameters applied to a trained machine learning model may be obtained from sensors and / or analyzers and / or from a user and / or user interface. In an embodiment, the GUI 500 may be displayed on to the user by an apparatus, a display, user devices, a computing device, a laptop, a tablet, a smartphone, or any device including a visual display and inputoutput connections. Alternatively, a user may interact with the DRA modeling system 102 using a separate apparatus, which may communicate with the DRA modeling system 102 via a communications network, such as, the internet, an intranet, a virtual private network, a local network, cellular data network, via hardwire, or other type of data connection. In an embodiment, the DRA modeling system 102 may generate such a GUI 500, for example, based on instructions stored in the memory 106 of the DRA modeling system 102. The DRA modeling system 102 may generate such a GUI 500 for an entire length of pipeline and / or for each section of the plurality of sections of the pipeline.

[0089] In an embodiment, GUI 500 may include fields enabling a user to enter values. For example, the GUI 500 may include a field for drag reduction desired 502 (for example, for a selected pipeline operation and / or to be entered as a decimal), a flow rate 504 (for example, for the fluid flowing within the pipeline and / or to be entered as barrels per hour(bph)), a grade code 506 (for example, displayed as a drop-down pre-populated with various types of hydrocarbons or, in another embodiment, as a blank field enabling the user to enter a grade code), a specific gravity (for example, of the hydrocarbon flowing through the pipeline), a viscosity (for example, to be entered as centistokes (cSt)), a temperature, and / or a distance (for example, for the length of the pipeline, a selected length of the pipeline, or the length of a section of the pipeline). One or more of the fields may include a tolerance band, for example, the flow rate 504 may include a tolerance band field enabling a user to enter an amount that the flow rate can vary between (for example, a flow rate of 15,700 bph may fluctuate by + / - 500 bph), the temperature may include a tolerance band field (for example, a temperature of 70 degrees Fahrenheit may vary by + / - 10 degrees Fahrenheit), and / or the distance may include a tolerance band field (for example, a distance of 40 miles may vary by + / -2 miles).

[0090] In an embodiment, the GUI 500 may include several buttons that when selected perform a function. The GUI 500 may include a fill grade properties button 516. When selected, the fill grade properties button 516 may automatically populate fields in the GUI 500 based on a selected grade code, for example, the specific gravity 508 and viscosity 510 may be populated. The GUI 500 may include a calculate button 518 that, when selected, may calculate an amount of DRA to inject into the pipeline, a length of the pipeline, or a section of the pipeline. The amount of DRA may be measured at PPM in relation to the amount of fluid flowing within the pipeline, the length of the pipeline, or the section of the pipeline. The GUI 500 may also include a calculate A & B with plot button 520 that when selected may determine the A and B coefficient of the DRA to be injected into the pipeline, the length of the pipeline, or the section of the pipeline.

[0091] Further, in another embodiment, selection of the calculate A & B with plot button 520 may generate a DRA curve 522, as illustrated in FIG. 5B. The DRA curve may display the A and B coefficient of the DRA. Further, the DRA curve 522 may display the amount of drag reduction experienced 524 versus the amount of DRA injected 526 (for example, measured as PPM) based on the fields entered into the GUI 500.

[0092] FIG. 6 is a flow diagram for training a machine learning model, according to an embodiment of the disclosure. Unless otherwise specified, the actions of method 600 may be completed within controller 402. Specifically, method 600 may be included in one or more programs, protocols, or instructions loaded into the memory of controller 402 and executed on the processor or one or more processors of the controller 402. In other embodiments, method 600 may be implemented in or included in components of FIGS. 1-3. The order in which the operations are described is not intended to be construed as a limitation, and any number of the described blocks may be combined in any order and / or in parallel to implement the methods.

[0093] At block 602, the controller 402 may determine, from a historical data set, a selected time where a selected fluid (for example, hydrocarbon liquid and / or renewable liquid) or distinct type of fluid flows through a section of a pipeline or through all or substantially all of the section of the pipeline, to ensure accurate training of the machine learning model. At block 604, the controller 402 may determine one or more parameters associated with the fluid flowing through the pipeline at the selected time. In an embodiment, the controller 402 may determine the one or more parameters based on a grade code associated with the fluid. In another embodiment, the historical data set may include the one or more parameters.

[0094] At block 606, the controller 402 may determine the amount and parameters of DRA injected into the pipeline at the selected time. The parameters associated with DRA, such as an A and B coefficient, may be indicated based on the DRA. At block 608, the controller 402 may determine one or more parameters associated with a pump operating at the selected time for the section of the pipeline. At block 610, the controller 402 may add each parameter over the selected time to a training data set.

[0095] For example, as shown in Table 1, various parameters may be included in or determined for a set of historical data. The historical data set may include a specific gravity (SG) for fluid within a section of the pipeline, a DRA parts per million (PPM) average for the section of pipeline with a selected linefill (in other words, fluid within the pipeline) at a selected moment, a DRA PPM standard deviation (dev) for a section of pipeline with the selected linefill at the selected moment, an actual drag reduction percentage, a viscosity average of the fluid within the section of the pipeline, an active temperature of the fluid within the section of the pipeline, a time stamp of the measurement of the data, a flow rate of the dev in the section of the pipeline, a flow rate average of the fluid within the section of the pipeline, a grade of the fluid within the pipeline, a mileage start and distance, a diameter of the section of the pipeline, and / or an identifier of the section of the pipeline. At selected times, the section of the pipeline may include a blend or mix of a fluid from a current pipeline operation and a fluid from a following pipeline operation. As noted, the portions of data including the blend or mix may be removed from the historical data, as shown in Table 1. Such a set of data may be, as described herein, may be utilized to train a machine learning model.Table 1

[0096] At block 612, the controller 402 may determine whether a selected threshold relating to an amount of data in a training set has been met and if not, the controller 402, at 5 block 612, may add additional data (in other words, moving back through method 600 and starting at block 602) to the training data set. If the selected threshold is met, the controller 402, at block 614, may preprocess the training data set. Preprocessing may include normalizing the data set and separating the data set into a first portion of the set data (also referred to as the training set) and a second portion of the data set (also referred to as the 10 testing set). For example, separation of the data may follow the 80 / 20 testing scheme. In such an embodiment, 80 percent of the data may be included in the first portion of the setdata, while 20 percent may be included in the second portion of the set data. Other schemes may be used, such as 90 / 10 or 70 / 30.

[0097] Once the training data set has been preprocessed, at block 616, the controller 402 may train the machine learning model with the first portion of data set (for example, the training set). At block 618, the controller 402 may test the machine learning model with the second portion of the data set. Testing the machine learning model may determine a performance indicator.

[0098] At block 620, if the performance indicator meets a second selected threshold, then the method 600 may end, resulting in or the output being a trained machine learning model. If the second selected threshold is not met, then the controller 402 may obtain additional data for training or retraining the machine learning model and / or may reuse different data within the training data set to retrain the machine learning model.

[0099] FIG. 7 is a flow diagram for using a trained machine learning model to determine an amount of drag reduction agent, according to an embodiment of the disclosure. Unless otherwise specified, the actions of method 700 may be completed within controller 402. Specifically, method 700 may be included in one or more programs, protocols, or instructions loaded into the memory of controller 402 and executed on the processor or one or more processors of the controller 402. In other embodiments, method 700 may be implemented in or included in components of FIGS. 1-3. The order in which the operations are described is not intended to be construed as a limitation, and any number of the described blocks may be combined in any order and / or in parallel to implement the methods.

[0100] At block 702, the controller 402 may obtain one or more parameters for a pipeline operation. The parameters may include data that corresponds to operation of pumps for each section of the pipeline, a DRA , and / or a hydrocarbon flowing through the pipeline. The parameters may include additional data points related to the DRA, the pipeline, and / or the properties of hydrocarbon flowing through the pipeline.

[0101] At block 704, the controller 402 may determine one or more of an amount of DRA for each section of the pipeline. In another embodiment, the controller 402 may determine an amount of drag reduction for each section of the pipeline. Such determinations may be based on application of the parameters to the trained machine learning model.

[0102] At block 706, if the amount of DRA is different than a previously set amount of DRA, then the controller 402, at block 708, may adjust the amount of DRA injected viasignal to a DRA injection device indicative of the amount of DRA that the DRA injection device should adjust to.

[0103] At block 714 then controller 402 may determine if the pipeline operation is complete. If the pipeline operation is not complete, the controller 402, at block 716, may obtain additional and / or new parameters for application to the trained machine learning mode. If the pipeline operation has complete, then the method 700 may end, as indicated at block 718.

[0104] In the drawings and specification, several embodiments of systems and methods to provide in-line mixing of hydrocarbon liquids have been disclosed, and although specific terms are employed, the terms are used in a descriptive sense only and not for purposes of limitation. Embodiments of systems and methods have been described in considerable detail with specific reference to the illustrated embodiments. However, it will be apparent that various modifications and changes may be made within the spirit and scope of the embodiments of systems and methods as described in the foregoing specification, and such modifications and changes are to be considered equivalents and part of this disclosure.

Claims

CLAIMSWHAT IS CLAIMED IS:

1. A method to determine an amount of drag reduction agent (DRA) to inject into one or more sections of pipeline, the method comprising: generating a training data set based on a plurality of data sets that each correspond to a distinct hydrocarbon flow for a selected section of the one or more sections of the pipeline; training a machine learning model with the training data set thereby to define a trained machine learning model; obtaining one or more parameters for a pipeline operation that correspond to (a) operation of one or more pumps for each of the one or more sections of the pipeline, (b) the DRA to be utilized for the pipeline operation, and (c) a hydrocarbon to flow through each of the one or more sections of the pipeline during the pipeline operation; determining, based on application of the one or more parameters to the trained machine learning model, an amount of the DRA to inject into each of the one or more sections of the pipeline at one or more selected times; and in response to determination of the amount of the DRA, injecting the amount of the DRA into the selected section of the one or more sections of the pipeline.

2. The method of claim 1, wherein the training data set includes one or more of properties that correspond to the hydrocarbon, a start time that the distinct hydrocarbon flow begins in for the selected section of the one or more sections of the pipeline, an end time that the distinct hydrocarbon flow begins in for the selected section of the one or more sections of the pipeline, a previous amount of the DRA injected into the selected section of the one or more sections of the pipeline, properties of the DRA injected into the selected section of the one or more sections of the pipeline, a pipeline diameter for the selected section of the one or more sections of pipeline, properties of the pump for the selected section of the one or more sections of the pipeline, a power amount used for the pump corresponding to the selected section of the one or more sections of the pipeline, a viscosity of the distinct hydrocarbon flow, a gravity of the distinct hydrocarbon flow, or a flow rate of the distinct hydrocarbon flow; and wherein a grade code defines the properties of the hydrocarbon.

3. The method of any of claims 1 or 2, wherein determining the amount of the DRA to inject into each of the one or more sections of the pipeline at the one or more selected times is further based on application of an amount of drag reduction input for the pipeline operation.

4. The method of any of claims 1-3, further comprising, prior to training the machine learning model comprising: determining whether the training data set includes an amount of data that exceeds a first selected threshold; and if the amount of data exceeds the first selected threshold, preprocessing the training data set.

5. The method of claim 4, wherein preprocessing the training data set comprises: normalizing the training data set thereby to define a normalized training data set; and splitting the training data set into a training set and a test set.

6. The method of claim 5, further comprising, subsequent to preprocessing the training data set: training the machine learning model with the training set; testing the machine learning model with the test set; determining if performance during testing of the trained machine learning model meets a second threshold; and if the performance does not meet the second threshold, re-training the machine learning model with a different portion of the training data set.

7. The method of any of claims 1-6, further comprising: determining, based on application of the one or more parameters to the trained machine learning model, an amount of drag reduction for the pipeline operation.

8. The method of claim 7, wherein determination of one or more of (i) an amount of DRA to inject into each of the one or more sections of the pipeline at one or moreselected times or (ii) an amount of drag reduction for the pipeline operation is further based on an input specifying an output of the trained machine learning model.

9. The method of any of claims 1-8, wherein the one or more parameters are obtained from one or more sensors or analyzers positioned at each of the one or more sections of the pipeline.

10. The method of any of claims 1-9, wherein training the trained machine learning model is performed for each section of the one or more sections of the pipeline and wherein determining an amount of the DRA to inject into each of the one or more sections of the pipeline at one or more selected times is for each section of the one or more sections of the pipeline and based on application of the one or more parameters corresponding to a selected section of the one or more sections of the pipeline to the trained machine learning model for the selected section of the one or more sections of the pipeline.

11. The method of any of claims 1-10, further comprising: receiving a plurality of parameters for a pipeline operation that correspond to (a) operation of one or more pumps for each of the one or more sections of the pipeline, (b) the DRA to be utilized for the pipeline operation, and (c) a t hydrocarbon to flow through each of the one or more sections of the pipeline during the pipeline operation, and wherein determining the amount of DRA to inject into each of the one or more sections of the pipeline at the one or more selected times is based on the plurality of parameters for the pipeline operation.

12. The method of claim 11, further comprising, during the pipeline operation: determining updated one or more parameters; determining, based on application of the updated one or more parameters and one or more of the plurality of parameters to the trained machine learning model, an updated amount of DRA to inject into each of the one or more sections of the pipeline at one or more selected times; and injecting the updated amount of DRA into the selected section of pipeline.

13. The method of claim 12, wherein determination of the updated amount of DRA occurs at selected time intervals.

14. The method of any of claims 11-13, wherein the plurality of parameters are received from one or more sensors or analyzers associated with one or more of (a) each of the one or more pumps, (b) each of one or more DRA injection devices, and (c) each of the one or more sections of the pipeline.

15. The method of any of claims 11-14, further comprising: determining parameters to include one or more of (a) properties of hydrocarbon flowing through the section of pipeline, (b) a current amount and properties of DRA injected into section of the pipeline, and (c) one or more of a speed or power of a pump for the section of pipeline; and determining one or more of (i) an adjusted amount of DRA to inject into the section of the pipeline or (ii) a drag reduction amount for the section of the pipeline based on application of the parameters to a trained machine learning model.

16. The method of claim 15, wherein the parameters are received from one or more sensors or analyzers associated with one or more of (a) each of the one or more pumps, (b) each of one or more DRA injection devices, and (c) each of the one or more sections of the pipeline.

17. The method of any of claims 15 or 16, wherein the drag reduction amount includes an adjusted amount of power for the pump of the section of the pipeline and the adjusted amount of DRA for the section of the pipeline.

18. The method of any of claims 15-17, wherein determination of the adjusted amount of DRA to inject into the section of the pipeline is further based on user input including entry of the drag reduction amount for the section of pipeline.

19. The method of any of claims 15-18, wherein determination of the parameters and the one or more of (i) the adjusted amount of DRA or (ii) the drag reduction amount occurs at selected time intervals during a pipeline operation.

20. The method of any of claims 15-19, wherein determination of the parameters or a portion of the parameters occurs in real-time at each of selected time intervals during pipeline operation.

21. The method of any of claims 1-20, wherein the parameters include a prior amount of DRA included in a hydrocarbon flow from a previous section of the pipeline to the section of the pipeline.

22. A method to determine an amount of drag reduction agent (DRA) to inject into one or more sections of pipeline, the method comprising: for each section of a plurality of sections of pipeline: performing the method of any of claims 1-21.

23. A method to determine one or more of an amount of drag reduction agent (DRA) to inject into a section of pipeline or a drag reduction amount for the section of pipeline, the method comprising: for each section of a plurality of sections of pipeline: obtaining one or more parameters for a pipeline operation that correspond to (a) operation of one or more pumps for a selected section of the plurality of sections of the pipeline, (b) the DRA to be utilized for the pipeline operation, or (c) a hydrocarbon to flow through the selected section of the plurality of sections of the pipeline during the pipeline operation; determining, based on application of the one or more parameters to a trained machine learning model for the selected section of the plurality of sections of the pipeline, one of (a) an amount of the DRA to inject into the selected section of the plurality of sections of the pipeline at one or more selected times or (b) an amount of drag reduction for the selected section of the plurality of sections of the pipeline; and one of:in response to determination of the amount of the DRA, injecting the amount of the DRA into the selected section of the plurality of sections of the pipeline, or in response to determination of the amount of the drag reduction, injecting a second amount of the DRA into the selected section of the plurality of sections of the pipeline.

24. The method of any of claims 1-23, wherein one or more parameters include flow rate, grade code, viscosity, gravity, or temperature of fluid flowing through the selected section of the plurality of sections of the pipeline or length of the selected section of the plurality of sections of the pipeline.

25. The method of any of claims 1-24, wherein one or more parameters include flow rate, gravity, or viscosity of fluid flowing through the selected section of the plurality of sections of the pipeline.

26. A computer readable medium comprising instructions executable to perform the method of any of claims 1-25.

27. A controller to determine an amount of drag reduction agent (DRA) to inject into one or more sections of pipeline, the controller comprising instructions executable to perform the method of any of claims 1-25.

28. A drag reduction agent (DRA) modeling circuitry including instructions executable to perform the method of any of claims 1-25.

29. A system to determine an amount of drag reduction agent (DRA) to inject into a hydrocarbon flow of one or more sections of pipeline, the system comprising: a plurality of pipeline sections; a plurality of pumps in fluid communication, each one of the plurality of pumps corresponding to one of the plurality of pipeline sections; a plurality of DRA containers to store DRA, each one of the plurality of DRA containers corresponding to one of the plurality of pipeline sections;a plurality of DRA injection devices, each one of the plurality of DRA injection devices to connect to one of the plurality of DRA containers and one of the plurality of pipeline sections, each one of the plurality of DRA injection devices configured to inject an amount of the DRA into each one of the plurality of pipeline sections, one or more sensors positioned to measure parameters at one or more of on, within, or proximal to one or more of (a) the plurality of pipeline sections, (b) the plurality of pumps, (c) the plurality of DRA containers, or (d) the plurality of DRA injection devices; and a DRA modeling circuitry including instructions executable to perform the method of any of claims 1-25.

30. A system to determine an amount of drag reduction agent (DRA) to inject into one or more sections of pipeline, the system comprising: one or more sensors positioned at one or more of on, within, or proximal to one or more components and configured to measure parameters associated with each of the one or more components including one or more of (a) each of the one or more sections of pipeline or (b) each of one or more pumps corresponding to one of the one or more sections of pipeline; and a DRA modeling circuitry configured to: obtain parameters measured by the one or more sensors, apply the parameters to a trained machine learning model, determine an adjusted amount of DRA to inject into each section of the pipeline based on application of the one or more parameters to the trained machine learning model, and initiate injection of the adjusted amount of DRA into each section of the pipeline.

31. The system of claim 30, further comprising DRA injection devices in signal communication with the DRA modeling circuitry and each one of the DRA injection devices corresponding to one of each section of the pipeline.

32. The system of claim 31, wherein initiation of injection of the adjusted amount of DRA for a section of the pipeline includes the DRA modeling circuitry communicatinga signal to a corresponding DRA injection device, the signal indicating the amount of DRA to inject and causing the corresponding DRA injection device to begin injection of the amount of DRA.

33. The system of claim 30, wherein the DRA modeling circuitry is further configured to obtain predetermined parameters for each section of the pipeline.

34. The system of claim 33, wherein determination of the adjusted amount of DRA to inject into each section of the pipeline is further based on application of the predetermined parameters to the trained machine learning model.

35. The system of claim 30, wherein the DRA modeling circuitry is further configured to obtain the parameters and determine the adjusted amount of DRA at preselected time intervals during a pipeline operation.

36. The system of claim 30, further comprising one or more pumps, each one positioned in fluid flow with each section of the pipeline, each of the one or more pumps configured to cause a fluid to flow through a section of the pipeline at a selected flow rate.

37. The system of claim 36, wherein the fluid comprises one or more of a hydrocarbonbased fluid or a renewable hydrocarbon-based fluid.

38. A system to determine an amount of drag reduction agent (DRA) to inject into a hydrocarbon flow of one or more sections of pipeline, the system comprising: a plurality of pipeline sections; a plurality of pumps in fluid communication, each one of the plurality of pumps corresponding to one of the plurality of pipeline sections; a plurality of DRA containers to store DRA, each one of the plurality of DRA containers corresponding to one of the plurality of pipeline sections; a plurality of DRA injection devices, each one of the plurality of DRA injection devices to connect to one of the plurality of DRA containers and one of the plurality of pipeline sections, each one of the plurality of DRA injection devices configured to inject an amount of the DRA into each one of the plurality of pipeline sections,one or more sensors positioned to measure parameters at one or more of on, within, or proximal to one or more of (a) the plurality of pipeline sections, (b) the plurality of pumps, (c) the plurality of DRA containers, or (d) the plurality of DRA injection devices; and a DRA modeling circuitry configured to: obtain the parameters measured by the one or more sensors, apply the parameters to a trained machine learning model, determine an adjusted amount of DRA to inject into each section of the pipeline based on application of the parameters to the trained machine learning model, and initiate injection of the adjusted amount of DRA into each section of the pipeline via a signal transmitted to one of the plurality of DRA injection devices and indicative of the adjusted amount of DRA.

39. The system of claim 38, wherein the parameters include one or more of (a) properties of hydrocarbon flowing through each of the plurality of pipeline sections, (b) a current amount and properties of DRA injected into each of the plurality of pipeline sections, (c) one or more of a speed or power of each of the plurality of pumps for each of the plurality of pipeline sections, or (d) an available amount of DRA in each of the plurality of DRA containers.

40. A controller to determine an amount of drag reduction agent (DRA) to inject into one or more sections of pipeline, the controller comprising: a first plurality of inputs each in signal communication with one or more of (a) one or more sensors configured to measure a first set of parameters or (b) a user interface configured to receive a second set of parameters, the controller configured to: apply one or more of the first set of parameters or the second set of parameters to a trained machine learning model, and determine an adjusted amount of DRA to inject into a selected section of pipeline based on application of the one or more of the first set of parameters or the second set of parameters to the trained machine learning model; anda first output in signal communication with a DRA injector device, the controller configured to cause the DRA injector device to inject the adjusted amount of DRA into the selected section of pipeline.

Images