ARTIFICIAL INTELLIGENCE BASED FRAMEWORK FOR THE OPTIMIZATION OF LITHIUM EXTRACTION IN RESERVOIR BRINE

Description

BACKGROUND

The extraction and production of oil and gas from a well, or an oil and gas field composed of at least one well, is a complex process. In general, optimization of an oil and gas field seeks to maximize hydrocarbon recovery while minimizing cost, where cost is accrued through the allocation of resources and energy.

A common byproduct of oil and gas production is water, often referred to as produced water or produced brine (or, sometimes simply brine). The composition of the produced water depends on the chemistry and thermophysical properties of the subsurface rocks with which the produced water has been in contact. The produced water may contain various alkali metals such as lithium, alkaline earth metals, and precipitated salts.

SUMMARY

This summary is provided to introduce a selection of concepts that are further described below in the detailed description. This summary is not intended to identify key or essential features of the claimed subject matter, nor is it intended to be used as an aid in limiting the scope of the claimed subject matter.

Embodiments disclosed herein generally relate to a method for joint and optimal production of hydrocarbons and lithium from a well, the method including obtaining field data for an oil and gas field with at least one well accessing at least one hydrocarbon reservoir. The method further includes obtaining a set of well operational parameters related to the oil and gas field and obtaining a set of lithium extraction configuration parameters related to the oil and gas field. The method further includes determining, with an artificial intelligence model, a predicted hydrocarbon production and a predicted lithium extraction from production fluids of the oil and gas field based on the field data, the set of well operational parameters, and the set of the lithium extraction configuration parameters and adjusting, automatically, the set of well operational parameters and the set of the lithium extraction configuration parameters to jointly optimize the predicted hydrocarbon production and the predicted lithium extraction.

Embodiments disclosed herein generally relate to a non-transitory computer-readable memory with computer-executable instructions stored thereon that, when executed on a processor, cause the processor to perform the following steps. The steps include obtaining field data for an oil and gas field with at least one well with access to at least one hydrocarbon reservoir. The field data includes a number of wells in the oil and gas field, well history data for each well in the oil and gas field, spatial data for each well in the oil and gas field, and water quality data. The steps further include obtaining a set of well operational parameters related to the oil and gas field and obtaining a set of lithium extraction configuration parameters related to the oil and gas field. The steps further include determining, with an artificial intelligence model, a predicted hydrocarbon production and a predicted lithium extraction from production fluids of the oil and gas field based on the field data, the set of well operational parameters, and the set of the lithium extraction configuration parameters. The steps further include adjusting, automatically, the set of well operational parameters and the set of the lithium extraction configuration parameters to jointly optimize the predicted hydrocarbon production and the predicted lithium extraction.

Embodiments disclosed herein generally relate to system including an oil and gas field with at least one well and at least one lithium extraction system, wherein operation of the at least one well is defined by a set of well operational parameters and operation and configuration of the at least one lithium extraction system is defined by a set of lithium extraction configuration parameters. The system further includes a plurality of field devices disposed throughout the oil and gas field, the plurality of field devices collecting field data for the oil and gas field. The field data includes a number of wells in the oil and gas field, well history data for each well in the oil and gas field, spatial data for each well in the oil and gas field, and water quality data. The system further includes a control system configured to adjust one or more of the field devices in the plurality of field devices and a computer. The computer is configured to obtain the field data for the oil and gas field, obtain the set of well operational parameters, obtain the set of lithium extraction configuration parameters, and determine, with an artificial intelligence model, a predicted hydrocarbon production and a predicted lithium extraction from production fluids of the oil and gas field based on the field data, the set of well operational parameters, and the set of the lithium extraction configuration parameters. The computer is further configured to adjust, automatically, the set of well operational parameters and the set of the lithium extraction configuration parameters to jointly optimize the predicted hydrocarbon production and the predicted lithium extraction.

Other aspects and advantages of the claimed subject matter will be apparent from the following description and the appended claims.

BRIEF DESCRIPTION OF DRAWINGS

FIG. 1 depicts a well environment in accordance with one or more embodiments.

FIG. 2 depicts a system in accordance with one or more embodiments.

FIG. 3A depicts a system in accordance with one or more embodiments.

FIG. 3B depicts a system in accordance with one or more embodiments.

FIG. 4 depicts a flowchart in accordance with one or more embodiments.

FIG. 5A depicts a flowchart in accordance with one or more embodiments.

FIG. 5B depicts a flowchart in accordance with one or more embodiments.

FIG. 6 depicts a neural network in accordance with one or more embodiments.

FIG. 7A depicts a recurrent neural network in accordance with one or more embodiments.

FIG. 7B depicts an unrolled recurrent neural network in accordance with one or more embodiments.

FIG. 7C depicts a long short-term memory network in accordance with one or more embodiments.

FIG. 8 depicts a system in accordance with one or more embodiments.

DETAILED DESCRIPTION

In general, embodiments of the disclosure include systems and methods for jointly optimizing lithium extraction and hydrocarbon production from an oil and gas field consisting of at least one production well. Herein, for simplicity, the material output of an oil and gas field, or that of a production well, is a produced fluid. The produced fluid may be multiphase and be composed of a variety of solid, liquid, and gaseous constituents. For example, the produced fluid may contain solid particulates like sand, mineral precipitates such as pipe scale, and corroded pipe, liquid such as water (referenced herein as “produced water”), hydrocarbons in both liquid (i.e., oil) and gas states (gaseous hydrocarbons may simply be referred to as “gas”), and other gases like carbon dioxide (CO₂) and hydrogen sulfide (H₂S).

In one or more embodiments, lithium is extracted from produced water acquired as a byproduct of hydrocarbon (i.e., oil and/or gas) production. The efficiency of a lithium extraction process and overall quantity lithium acquired are affected by a set of lithium extraction configuration parameters that define a lithium extraction system and its operation. Likewise, the operation of an oil and gas field is directly controlled through assignation of a set of well operational parameters to a given state or value. In general, an interaction exists between the set of lithium extraction configuration parameters, the set of well operational parameters, and the production and extraction rates of hydrocarbons and lithium, respectively. In one or more embodiments, the set of lithium extraction configuration parameters and the set of well operational parameters are adjusted, automatically and in real time, to simultaneously maximize both hydrocarbon production and lithium extraction from the produced fluid(s) of an oil and gas field.

In one or more embodiments, hydrocarbon production and lithium extraction from an oil and gas field, or a production well of the oil and gas field, are predicted using an artificial intelligence (AI) model based on (or accepts as inputs), at least, the set of lithium extraction configuration parameters and the set of well operational parameters. In one or more embodiments, the AI model may further make use of well history data (e.g., the historical production output of at least one production well) and a description of the produced water (e.g., lithium concentration). Using the AI model to predict hydrocarbon production and lithium extraction, an optimal set of both well operational parameters and lithium extraction configuration parameter may be determined using an optimization wrapper.

FIG. 1 shows a schematic diagram in accordance with one or more embodiments. More specifically, FIG. 1 illustrates a well environment (100) that includes a hydrocarbon reservoir (“reservoir”) (102) located in a subsurface formation (“formation”) (104) and a well system (106). The formation (104) may include a porous formation that resides underground, beneath the Earth's surface (“surface”) (108). In the case of the well system (106) being a hydrocarbon well, the reservoir (102) may include a portion of the formation (104). The formation (104) and the reservoir (102) may include different layers (referred to as subterranean intervals or geological intervals) of rock having varying characteristics, such as varying degrees of permeability, porosity, capillary pressure, and resistivity. In other words, a subterranean interval is a layer of rock having approximately consistent permeability, porosity, capillary pressure, resistivity, and/or other characteristics. For example, the reservoir (102) may be an unconventional reservoir or tight reservoir in which fractured horizontal wells are used for hydrocarbon production. In the case of the well system (106) being operated as a production well, the well system (106) may facilitate the extraction of hydrocarbons (or “hydrocarbon production,” or simply “production” when appropriate based on context) from the reservoir (102).

In some embodiments, the well system (106) includes a wellbore (120), a well sub-surface system (122), a well surface system (124), and a well control system (e.g., SCADA system (126)). The control system may control various operations of the well system (106), such as well production operations, well completion operations, well maintenance operations, and reservoir monitoring, assessment and development operations. In some embodiments, the control system includes a computer system that is the same as or similar to that of computer system depicted in FIG. 6 with its accompanying description.

The wellbore (120) may include a bored hole that extends from the surface (108) into a target zone (i.e., a subterranean interval) of the formation (104), such as the reservoir (102). An upper end of the wellbore (120), terminating at or near the surface (108), may be referred to as the “up-hole” end of the wellbore (120), and a lower end of the wellbore, terminating in the formation (104), may be referred to as the “down-hole” end of the wellbore (120). The wellbore (120) may facilitate the circulation of drilling fluids during drilling operations, the flow the produced fluid (121) (e.g., hydrocarbons, water, etc.) from the subsurface to the surface (108) during production operations, the injection of substances (e.g., water) into the formation (104) or the reservoir (102) during injection operations, or the communication of monitoring devices (e.g., logging tools) into the formation (104) or the reservoir (102) during monitoring operations (e.g., during in situ logging operations). For example, the logging tools may include logging-while-drilling tool or logging-while-tripping tool for obtaining downhole logs.

In some embodiments, the well sub-surface system (122) includes casing installed in the wellbore (120). For example, the wellbore (120) may have a cased portion and an uncased (or “open-hole”) portion. The cased portion may include a portion of the wellbore having casing (e.g., casing pipe and casing cement) disposed therein. The uncased portion may include a portion of the wellbore not having casing disposed therein. In embodiments having a casing, the casing defines a central passage that provides a conduit for the transport of tools and substances through the wellbore (120). For example, the central passage may provide a conduit for lowering logging tools into the wellbore (120), a conduit for the flow of production (121) (e.g., oil and gas) from the reservoir (102) to the surface (108), or a conduit for the flow of injection substances (e.g., water) from the surface (108) into the formation (104). In some embodiments, the well sub-surface system (122) includes production tubing installed in the wellbore (120). The production tubing may provide a conduit for the transport of tools and substances through the wellbore (120). The production tubing may, for example, be disposed inside casing. In such an embodiment, the production tubing may provide a conduit for some or all of the produced fluid (121) (e.g., oil, gas, water) passing through the wellbore (120) and the casing.

In some embodiments, the well sub-surface system (122) further includes various control components and sensors disposed down-hole. For example, in one or more embodiments, the well sub-surface system (122) includes an inflow control valve (ICV). An ICV is an active component usually installed during well completion. The ICV may partially or completely choke flow into a well. Generally, multiple ICVs may be installed along the reservoir section of a wellbore. Each ICV is separated from the next by a packer. Each ICV can be adjusted and controlled to alter flow within the well and, as the reservoir depletes, prevent unwanted fluids from entering the wellbore. The well sub-surface system (122) may further include a subsurface safety valve (SSSV). The SSSV is designed to close and completely stop flow in the event of an emergency. Generally, an SSSV is designed to close on failure. That is, the SSSV requires a signal to stay open and loss of the signal results in the closing of the valve. The well sub-surface system (122) can further include a permanent downhole monitoring system (PDHMS) (170). The PDHMS (170) consists of a plurality of sensors, gauges, and controllers to monitor subsurface flowing and shut-in pressures and temperatures. As such, a PDHMS (170) may indicate, in real-time, the state or operating condition of subsurface equipment and the fluid flow.

In some embodiments, the well surface system (124) includes a wellhead (130). The wellhead (130) may include a rigid structure installed at the “up-hole” end of the wellbore (120), at or near where the wellbore (120) terminates at the Earth's surface (108). The wellhead (130) may include structures (called “wellhead casing hanger” for casing and “tubing hanger” for production tubing) for supporting (or “hanging”) casing and production tubing extending into the wellbore (120). Produced fluid (121) may flow through the wellhead (130), after exiting the wellbore (120) and the well sub-surface system (122), including, for example, the casing and the production tubing. In some embodiments, the well surface system (124) includes flow regulating devices that are operable to control the flow of substances into and out of the wellbore (120). For example, the well surface system (124) may include one or more production valves (132) that are operable to control the flow of the produced fluid (121). For example, a production valve (132) may be fully opened to enable unrestricted flow of the produced fluid (121) from the wellbore (120), the production valve (132) may be partially opened to partially restrict (or “throttle”) the flow of the produced fluid (121) from the wellbore (120), and production valve (132) may be fully closed to fully restrict (or “block”) the flow of produced fluid (121) from the wellbore (120), and through the well surface system (124).

In some embodiments, the wellhead (130) includes a choke assembly. For example, the choke assembly may include hardware with functionality for opening and closing the fluid flow through pipes in the well system (106). Likewise, the choke assembly may include a pipe manifold that may lower the pressure of fluid traversing the wellhead. As such, the choke assembly may include a set of high pressure valves and at least two chokes. These chokes may be fixed or adjustable or a mix of both. Redundancy may be provided so that if one choke has to be taken out of service, the flow can be directed through another choke. In some embodiments, pressure valves and chokes are communicatively coupled to the well control system (e.g., SCADA system (126)).

Keeping with FIG. 1, in some embodiments, the well surface system (124) includes a surface sensing system (134). The surface sensing system (134) may include sensors for sensing characteristics of substances, including the produced fluid (121), passing through or otherwise located in the well surface system (124). The characteristics may include, for example, pressure, temperature and flow rate of produced fluid (121) flowing through the wellhead (130), or other conduits of the well surface system (124), after exiting the wellbore (120).

In one or more embodiments, the well surface system (124) includes a multiphase flow meter (MPFM). The MPFM monitors the flow rate of the produced fluid (121) by constituent. That is, the MPFM may detect the instantaneous amount of gas, oil, and water. As such, the MPFM indicates percent water cut (% WC) and the gas-to-oil ratio (GOR). Additionally, the MPFM (123) may measure pressure and fluid density.

In some embodiments, the surface sensing system (134) includes a surface pressure sensor (136) operable to sense the pressure of the produced fluid (121) flowing through the well surface system (124), after it exits the wellbore (120). The surface pressure sensor (136) may include, for example, a wellhead pressure sensor that senses a pressure of the produced fluid (121) flowing through or otherwise located in the wellhead (130). In some embodiments, the surface sensing system (134) includes a surface temperature sensor (138) operable to sense the temperature of the produced fluid (121) flowing through the well surface system (124), after it exits the wellbore (120). The surface temperature sensor (138) may include, for example, a wellhead temperature sensor that senses a temperature of the produced fluid (121) flowing through or otherwise located in the wellhead (130), referred to as “wellhead temperature” (T_wh). In some embodiments, the surface sensing system (134) includes a flow rate sensor (139) operable to sense the flow rate of the produced fluid (121) flowing through the well surface system (124), after it exits the wellbore (120). The flow rate sensor (139) may include hardware that senses a flow rate of the produced fluid (121) (Q_wh) passing through the wellhead (130).

In accordance with one or more embodiments, during operation of the well system (106), the control system (e.g., SCADA system (126)) collects and records well data (140) for the well system (106). The well data (140) may include, for example, a record of measurements of wellhead pressure (P_wh) (e.g., including flowing wellhead pressure), wellhead temperature (T_wh) (e.g., including flowing wellhead temperature), wellhead volume flow rate (Q_wh) over some or all of the life of the well (106), and water cut data. The well data (140) may further include wellhead data regarding the choke assembly and data referring to the states of subsurface valve(s) (e.g., ICV), if any, and other sensor data collected and received by the PDHMS (170).

In some embodiments, the measurements are recorded in real-time, and are available for review or use within seconds, minutes, or hours of the condition being sensed (e.g., the measurements are available within 1 hour of the condition being sensed). In such an embodiment, the wellhead data (140) may be referred to as “real-time” well data (140). Real-time well data (140) may enable an operator of the well (106) to assess a relatively current state of the well system (106), and make real-time decisions regarding development of the well system (106) and the reservoir (102), such as on-demand adjustments in regulation of the produced fluid (121) from the well.

The various valves, pressure gauges and transducers, sensors, and flow meters depicted of a well may be considered devices of an oil and gas field. As described, these devices may be disposed both above and below the surface of the Earth. These devices are used to monitor and control components and sub-processes of an oil and gas field. It is emphasized that the plurality of oil and gas field devices described in reference to FIG. 1 are non-exhaustive. Additional devices, such as electrical submersible pumps (ESPs) (not shown) may be present in an oil and gas field with their associated sensing and control capabilities. For example, an ESP may monitor the temperature and pressure of a fluid local to the ESP and may be controlled through adjustments to ESP speed or frequency.

The plurality of oil and gas field devices may be distributed, local to the sub-processes and associated components, global, connected, etc. The devices may be of various control types, such as a programmable logic controller (PLC) or a remote terminal unit (RTU). For example, a programmable logic controller (PLC) may control valve states, pipe pressures, warning alarms, and/or pressure releases throughout the oil and gas field. In particular, a programmable logic controller (PLC) may be a ruggedized computer system with functionality to withstand vibrations, extreme temperatures, wet conditions, and/or dusty conditions, for example, around a well system (106). With respect to an RTU, an RTU may include hardware and/or software, such as a microprocessor, that connects sensors and/or actuators using network connections to perform various processes in the automation system. As such, a distributed control system may include various autonomous controllers (such as remote terminal units) positioned at different locations throughout the oil and gas field to manage operations and monitor sub-processes. Likewise, a distributed control system may include no single centralized computer for managing control loops and other operations.

In accordance with one or more embodiments, FIG. 1 depicts a supervisory control and data acquisition (SCADA) system (126). A SCADA system (125) is a control system that includes functionality for device monitoring, data collection, and issuing of device commands. The SCADA system (126) enables local control at an oil and gas field as well as remote control from a control room or operations center. To emphasize that the SCADA system (126) may monitor and control the various devices of an oil and gas field, dashed lines connecting some of the oil and gas field devices to the SCADA system (125) are shown in FIG. 1.

In review, and in accordance with one or more embodiments, a plurality of field devices are disposed throughout a well system (106). A field device may be disposed below the surface (108) and considered part of the well subsurface system (122) (e.g., a component of the PDHMS (170)) or located above the surface (108) and considered part of the well surface system (124). Generally, field devices can measure or sense a property, control a state or process of the well system (106), or provide both sensory and control functionalities. For example, a state of a valve may include an indication of whether the valve is open or closed. In some instances, the state of a valve may be given by some percentage of openness (or closedness). As such, a field device, which may be the valve itself, can determine and transmit the state of the valve and therefore act as a sensor or sensory device. Further, a field device, which may be the valve itself, can alter or change the state of the valve by receiving a signal from the SCADA system (126). Sensed or measured properties of the well system (106) are stored and/or collected as well data (140) for the well, regardless if the sensed or measured property was determined by a field device of the well sub-surface system (122) or the well surface system (124).

Operation of the well system (106) may be controlled or dictated through one or more well control parameters (145). The well control parameters (145) may represent and/or prescribe an operational state of the devices of the well system. Thus, in one or more embodiments, the well system (106) is controlled through a control system (e.g., SCADA system (126)) that determines and transmits a command signal to the field devices of the well system (106) according to the well control parameters (145).

In one or more embodiments, the well system (106) may be an injection well. The injection well injects, or places, a fluid into porous subsurface formations such as a reservoir. The injected fluid may be composed of brine, freshwater, steam, polymers, carbon dioxide, and other chemical agents. The injected fluid may be tailored to the subsurface formations and further account for the location of one or more production wells in an oil and gas field to displace and aid in the extraction of oil and gas. Accordingly, well control parameters (145) for an injection well may include the composition of the injection fluid and its volume flow rate into the subsurface. When an oil and gas field is composed of more than one injection wells, the well control parameters (145) may further dictate a pattern injection strategy for the oil and gas field.

In one or more embodiments, the well system (106) may be a hydraulic fracturing well. In hydraulic fracturing, water, sand, and/or other chemicals may be injected into a well to break up underground bedrock and improve accessibility to oil and gas reserves. Again, the operation of a hydraulic fracturing well, as well as the composition of a fracturing fluid the processes driving its injection into the subsurface, are defined and controlled by well control parameters (145) for the hydraulic fracturing well.

Thus, regardless of the type of well system (106), the operation of the well system (106) and specification of materials and processes associated with the operation well system (106) are encompassed by the well control parameters (145) for the well system (106). The well control parameters (145) are both monitored and controlled by a control system (e.g., SCADA system (126)). The control system need not be proximate the well system (106) but may be located at a remote location relative to the well system (106).

In accordance with one or more embodiments, the well system (106) may be associated with or otherwise include a lithium extraction system (160). Lithium and its compounds are widely used in manufactured glass, ceramics, greases, batteries, refrigerants, chemical reagents, and other industries. Lithium demand is expected to grow continuously and dramatically in the coming years as different types of lithium batteries are the most promising candidates for powering electric or hybrid vehicles. Lithium can be present in produced water during the operation of a production well. The concentration of lithium in produced water is generally in the range of 100 to 1,000 mg/L. Because produced water is generally considered a waste byproduct in the pursuit of hydrocarbon extraction and production, the produced water represents an underutilized lithium resource.

In accordance with one or more embodiments, the lithium extraction system (160) removes lithium from the produced fluid (121) of a well; specifically, the lithium is removed from the produced water. In one or more embodiments, the lithium extraction system (160) uses evaporative methods to extract lithium from the produced water. In one or more embodiments, the lithium extraction system (160) uses ion exchange adsorption methods based on lithium-ion sieves to extract lithium from the produced water. An ion exchange adsorption method to remove lithium from the produced water will use at least one lithium adsorbent material, such as a mineral oxide, clay mineral, silicotitanates, or zirconium phosphate material. The types of lithium adsorbent materials and their sequence (if more than one is used) is specified using configuration parameters (164) of the lithium extraction system. In addition to specifying the configuration of adsorbent materials, the configuration parameters (164) of the lithium extraction system (160), similar to the well control parameters (145) of a well system (106), define and prescribe the operations of the lithium extraction system (160). As will be described later, determination of the configuration parameters (164) that result in optimal lithium extraction from produced water can depend on properties of the produced water (stored as water quality data (162)) and the operation of the well system (106) (i.e., the well control parameters (145)).

In accordance with one or more embodiments, one or more geochemical analysis techniques are applied, at least periodically, to the produced water of a well system (106). The results of the geochemical analysis techniques are stored as water quality data (162). The water quality data (162) indicates the concentration of, at least, lithium in the produced water. In some embodiments, the water quality data (162) may further indicate the presence and relative quantity (i.e., concentration) of other elements or compounds in the produced water such as potassium, sodium, boron, and precipitated salts. In one or more embodiments, the geochemical analysis techniques further result in water quality data (162) indicative of the ratio of alkali metals and alkaline earth metals to lithium and other phase chemistry information.

An oil and gas field may include more than one well. In instances where an oil and gas field has more than one well, the well data (140) may be described collectively as field data. In addition to the well data (140) of each well, the field data can include the number of wells in the oil and gas field as well as the location of each well and spatial distances between pairs of wells. For example, when the location of wells on the surface of the Earth are given as longitude and latitude coordinates, the spatial distance between any pair of wells may be computed using the Haversine formula. Further, the field data can include the wellbore geometry for each well along with data collected while drilling and/or completed each well (e.g., petrophysical information). Note that an oil and gas field may contain only one well.

Similarly, and oil and gas field may include more than one lithium extraction system (160). For example, an oil and gas field may have individual and independent lithium extraction system for each of its production wells. Consequently, under such an arrangement, the configuration parameters (164) of each lithium extraction system may be unique and/or different dependent on the produced fluid coming from their well systems as represented by the water quality data (162) of each lithium extraction system. In other cases, a single lithium extraction system (160) is applied to a comingled mixture of produced fluids (or produced water streams) originating from two or more well systems of an oil and gas field.

Considering an oil and gas field with one or more wells and one or more lithium extraction systems, the well data (140) of each well and water quality data (162) of each lithium extraction system may be grouped together and collectively referred to a field data. Thus, field data for an oil and gas field will contain information about the number of wells in the oil and gas field and will include well history data for each well in the oil and gas field. The well history data includes the well data (140) for a given well (e.g., production rates, percent water cut, operating conditions such as temperature and pressure at surface and subsurface locations, etc.). The field data also includes spatial data for each well in the oil and gas field (e.g., location of each well and spatial distance between wells) and all water quality data.

For an oil and gas field with at least one well, the well control parameters (145) for each well may be collectively referred to as a set of well operational parameters. That is, the set of well operational parameters includes the well control parameters for every well of an oil and gas field. Likewise, when the oil and gas field has at least one lithium extraction system, the configuration parameters (164) for each lithium extraction system may be collectively referenced as a set of lithium extraction configuration parameters.

Oil and gas field devices, like those shown in FIG. 1 (and others not shown), monitor and govern the behavior of the components and sub-processes of the oil and gas field. Therefore, the productivity of the oil and gas field is directly affected, and may be altered, by the devices. Generally, complex interactions between oil and gas field components and sub-process exist such that configuring field devices for optimal production is a difficult and laborious task. Further, the state and behavior of oil and gas fields is transient over the lifetime of the constituent wells requiring continual changes to the field devices to enhance production. Changes to field devices are further complicated when considering not only hydrocarbon production but lithium extraction from produced water of the oil and gas field. Changes to field devices of the oil and gas field and changes to the configuration of one or more lithium extraction systems of the oil and gas field may be required to optimize lithium extraction from the oil and gas field. In general, changes to field devices of an oil and gas field may not lead to both optimal hydrocarbon production and lithium extraction.

In one aspect, embodiments disclosed herein relate to a system for determining the set of well operational parameters and the set of lithium extraction configuration parameters that simultaneously optimize (at least with respect to an optimization weighting factor) hydrocarbon production and lithium extraction from an oil and gas field. The optimal set of well operational parameters and lithium extraction configuration parameters are determined with an artificial intelligence model taking into consideration the current state of the oil and gas field as monitored by the plurality of oil and gas field devices and further in view of the relative location of the one or more wells of the oil and gas field, petrophysical information, and water quality data. In accordance with one or more embodiments, the set of well operational parameters and the set of lithium extraction configuration parameters may be adjusted automatically, and in real-time, through a control system, such as the SCADA system (126).

In accordance with one or more embodiments, field data from the oil and gas field are processed with an artificial intelligence model to determine the optimal set of operational well parameters and lithium extraction configuration parameters for the oil and gas field. Artificial intelligence, broadly defined, is the extraction of patterns and insights from data. The phrases “artificial intelligence”, “machine learning”, “deep learning”, and “pattern recognition” are often convoluted, interchanged, and used synonymously throughout the literature. This ambiguity arises because the field of “extracting patterns and insights from data” was developed simultaneously and disjointedly among a number of classical arts like mathematics, statistics, and computer science. For consistency, the term artificial intelligence (AI), will be adopted herein, however, one skilled in the art will recognize that the concepts and methods detailed hereafter are not limited by this choice of nomenclature.

Artificial intelligence (AI) model types may include, but are not limited to, neural networks, random forests, generalized linear models, and Bayesian regression. Further, as defined herein, AI may include algorithmic search methods and optimization methods such as a line search or the genetic algorithm. AI model types are usually associated with additional “hyperparameters” which further describe the model. For example, hyperparameters providing further detail about a neural network may include, but are not limited to, the number of layers in the neural network, choice of activation functions, inclusion of batch normalization layers, and regularization strength. The selection of hyperparameters surrounding a model is referred to as selecting the model “architecture.” Generally, multiple model types and associated hyperparameters are tested and the model type and hyperparameters that yield the greatest predictive performance on a hold-out set of data is selected.

As noted, the objective of the AI model is to determine the optimal set of well operational parameters and optimal set of lithium extraction configuration parameters for an oil and gas field. In accordance with one or more embodiments, FIG. 2 depicts the interactions between the AI model, field data, set of operational well parameters, and the set of lithium extraction configuration parameters.

FIG. 2 depicts an oil and gas field (e.g., oil and gas field Z (202)) that includes a production well (e.g., production well A (252)), an injection well (e.g., injection well B (255)), and a lithium extraction system (e.g., lithium extraction system C (258)). It is noted that embodiments of the instant disclosure are not limited to oil and gas fields as depicted in FIG. 2. As seen in FIG. 2, the production well (e.g., production well A (252)) has associated with itself well data (e.g., well data A (253) and well control parameters (e.g., well control parameters A (254)). Likewise, the injection well (e.g., injection well B (255)) has associated with itself well data (e.g., well data B (256)) and well control parameters (e.g., well control parameters B (255)). The lithium extraction system (e.g., lithium extraction system C (258)) has associated with itself configuration parameters (e.g., configuration parameters C (259)) and water quality data (e.g., water quality data C (260)). The water quality data may originate from samples acquired at different times and at different locations over the oil and gas field (e.g., oil and gas field Z (202)).

As seen, oil and gas field device data are collected from the plurality of devices of the oil and gas field and are stored as well history data (e.g., well history data G (262)). That is, the well history data contains all field device data of the oil and gas field. The well history data is alongside other data items, described below, as field data (e.g., field data D (260)). The device data, and thus the well history data (e.g., well history data G (262)) may include measurements of temperature, pressure, percent water cut (% WC), and gas-to-oil ratio (GOR) from one or more field devices disposed throughout the oil and gas field. Likewise, subsurface measurements, such as temperature and pressure, may be collected and received from well subsurface systems. The device data may further include frequency, speed, pressure, and temperature measurements from one or more electrical submersible pumps (ESPs), pressure readings from a plurality of pressure transducers, and pressure, temperature, and valve states. Additionally, the device data includes the current settings of the ICVs and choke valves of the oil and gas field, if present. In one or more embodiments, the well history data (e.g., well history data G (262)) further includes the production history and injection history of the production and injection wells of the oil and gas field, respectively, if present. In short, the well history data (e.g., well history data G (262)) includes, at least, the well data (e.g., well data A (253), well data B (256)) for every well of the oil and gas field. One with ordinary skill in the art will appreciate that additional field devices may be employed in an oil and gas field and that additional associated well history data may be collected without departing from the scope of this disclosure.

In accordance with one or more embodiments, field data (e.g., field data D (260)) further includes information regarding the number of wells (and their types) of the oil and gas field (e.g., number of wells E (264)) as well as the location of the wells of the oil and gas field including distances between wells as spatial data (e.g., spatial data F (266)). In one or more embodiments, the field data (e.g., field data D (260)) further includes water quality data for the entire field (e.g., field water quality data (268)) that includes all water quality data associated with any number of lithium extraction systems (e.g., lithium extraction system C (258)) of the oil and gas field (e.g., water quality data C (260)). The water quality data of the field can include geochemical analyses of water samples acquired across the oil and gas field (e.g., from the wells). In one or more embodiments, the water quality data includes, at least, an indication of the concentration of lithium present in a produced water stream.

As previously discussed, for an oil and gas field with at least one well, the well control parameters for each well (e.g., well control parameters A (254), well control parameters B (257)) may be collectively referred to as a set of well operational parameters (e.g., set of well operational parameters H (270)). That is, the set of well operational parameters includes the well control parameters for every well of an oil and gas field. Likewise, when the oil and gas field has at least one lithium extraction system, the configuration parameters for each lithium extraction system (e.g., configuration parameters C (259)) may be collectively referenced as a set of lithium extraction configuration parameters (e.g., set of lithium extraction configuration parameters J (280)).

Continuing with FIG. 2, and in accordance with one or more embodiments, the field data (e.g., field data D 260), set of well operational parameters (e.g., set of well operational parameters H (270)), and the set of lithium extraction configuration parameters (e.g., set of lithium extraction configuration parameters J (280)) are processed by an AI model (e.g., AI Model M (206)). In one or more embodiments, the result of the AI model is a prediction of the hydrocarbons that will be produced (e.g., predicted hydrocarbon production P (207)) and a prediction of the lithium that will be extracted (e.g., predicted lithium extraction Q (208)) from the oil and gas field operation according to the set of well operational parameters (e.g., set of well operational parameters H (270)), the set of lithium extraction parameters (e.g., set of lithium extraction configuration parameters J (280)), and in view of the field data (e.g., field data D (260)). The predictions for hydrocarbon production and lithium extraction can be given as rates, quantities over a given time period, or both. While FIG. 2 depicts the AI model as being applied to an oil and gas field, herein, the minimum number of wells and lithium extraction systems required to form an oil and gas field are each one. As such, in one or more embodiments, the AI model is applied to a single well and single lithium extraction system. In one or more embodiments, the AI model is applied to a single well and a single lithium extraction system even when more than one well or more than one lithium extraction system is included in the oil and gas field. That is, the AI model may be applied across an oil and gas field or across individual wells and lithium extraction systems of the oil and gas field without limitation.

In accordance with one or more embodiments, the device data, the set of well operational parameters, and the set of lithium extraction parameters may be pre-processed before being processed by AI model. Pre-processing may include activities such as, numericalization, filtering and/or smoothing of the data, scaling (e.g., normalization) of the data, feature selection, outlier removal (e.g., z-outlier filtering) and feature engineering. Feature selection comprises identifying and selecting a subset of field data with the greatest discriminative power with respect to predicting the hydrocarbon production and lithium extraction. For example, in one embodiment, discriminative power may be quantified by calculating the strength of correlation between elements of the field data and the predicted quantities. Consequently, in some embodiments, not all of the field data need be passed to the AI model. Feature engineering encompasses combining, or processing, various field data to create derived quantities. The derived quantities can be processed by the AI model. For example, the field data may be processed by one or more “basis” functions such as a polynomial basis function or a radial basis function. In some embodiments, the field data is passed to the AI model without pre-processing. Many additional pre-processing techniques exist such that one with ordinary skill in the art would not interpret those listed here as a limitation on the present disclosure.

In accordance with one or more embodiments, the predicted hydrocarbon production (e.g., predicted hydrocarbon production P (207)) and predicted lithium extraction (e.g., predicted lithium extraction Q (208)) are used to determine the set of well operational parameters (e.g., set of well operational parameters H (270)) and the set of lithium extraction configuration parameters (e.g., set of lithium extraction configuration parameters J (280)) that simultaneously optimize the hydrocarbon production and lithium extraction. That is, the set of well operational parameters and the set of lithium extraction configuration parameters define an operational state of an oil and gas field and thus affect the output of the oil and gas field, namely, hydrocarbon production and lithium extraction. Upon identifying the set of well operational parameters and the set of lithium extraction configuration parameters that jointly optimize hydrocarbon production and lithium extraction, these sets of parameters may be applied across the oil and gas field automatically using a control system (e.g., SCADA system (126)). In one or more embodiments, hydrocarbon production and lithium extraction, as monitored by at least one device from the plurality of oil and gas field devices, and collected as field data, are continuously monitored to ensure that the accepted set of well operational parameters and set of lithium extraction configuration parameters maintain the oil and gas field at optimal hydrocarbon production and lithium extraction.

FIG. 3A depicts an embodiment of using collected device data to determine the set of well operational parameters and the set of lithium extraction configuration parameters that optimize both hydrocarbon production and lithium extraction. In one or more embodiments, the AI model is trained using previously acquired, or historic, modelling data. The modelling data is partitioned into “inputs” (302) and “outputs” (304). The outputs include device data representative of oil and gas production and lithium extraction. The inputs include the set of well operational parameters, the set of lithium extraction configuration parameters, and all other field data retained, or derived quantities created, after pre-processing (if applied). The field data may be collected over a period of time or may be acquired from analogous oil and gas fields. The modelling data, partitioned into inputs (302) and outputs (304), are used to train an AI model (306). The trained AI model (e.g., trained AI model (305)) may be of any type known in the art. In some embodiments, multiple AI model types and/or architectures may be used. Generally, the AI model type and architecture with the greatest performance on a set of hold-out data is selected. Greater detail surrounding the training procedure for an AI model will be provided below in the context of a neural network and long short-term memory (LSTM) network. However, generally, training an AI model involves processing data to develop a functional relationship between elements of the data. The result of the training procedure is a trained AI model (e.g., trained AI Model A (305)). The trained AI model may be described as a function relating the inputs (302) and the outputs (304). That is, the AI model may be mathematically represented as outputs=ƒ (inputs), such that given an input (302) the AI model (305) may produce an output (304). In one or more embodiments, the trained AI model, upon processing an input, produces two outputs, namely a predicted hydrocarbon production (P) and a predicted lithium extraction (E). With a trained AI model, an optimization wrapper (depicted as Block 308) is used to invert the model to determine the set of well operational parameter and the set of lithium extraction configuration parameters that optimize hydrocarbon production and lithium extraction. Mathematically, the optimization takes the form

arg ⁢ max S 1 , S 2 ⁢ P + α ⁢ E subject ⁢ to : device ⁢ constraints , ( 1 )

where the quantities P and E, representing the predicted hydrocarbon production and the predicted lithium extraction, respectively, are determined using the trained AI model. Further, in EQ. 1, the set of well operational parameters is denoted as S₁ and the set of lithium extraction configuration parameters is denoted as S₂. Thus, the optimization wrapper (308) maximizes the predicted hydrocarbon production and the predicted lithium over the set of well operational parameters and the set of lithium extraction configuration parameters. EQ. 1 further makes uses of an optimization weighting factor (a). In EQ. 1, the optimization weighting factor is applied to the predicted lithium extraction, however, this need not be the case. In some embodiments, the optimization weighting factor is applied to (as a product) to the predicted hydrocarbon production. The optimization weighting factor serves, at least, two purposes. First, the optimization weighting factor acts to scale either the predicted hydrocarbon production or the predicted lithium extraction (as shown in EQ. 1). Second, the optimization weighting factor weights either the predicted lithium extraction or the predicted hydrocarbon production relative to the other output. In one or more embodiments, the optimization weighing factor is a predefined scalar such that the optimization wrapper (308), when applied to a trained AI model processing field data and parameterized by the set of well operational parameters and the set of lithium extraction configuration parameters, returns a single and optimal set of well operational parameters and a single and optimal set of lithium extraction configuration parameters. In other embodiments, the optimization weighting factor is an array such that the optimization wrapper (308) returns an array of optimal sets of well operational parameters and an array of optimal sets of lithium extraction configuration parameters. In this case, the optimization weighting factor array and associated sets of well operational parameters and lithium extraction configuration parameters define a so-called Pareto front. In such a case, an individual set of well operational parameters and set of lithium configuration parameters can be selected by a subject matter expert or automatically using a predefined criterion. One with ordinary skill in the art will appreciate that maximization and minimization may be made equivalent through simple techniques such as negation. As such, the choice to represent the optimization as a maximization as shown in EQ. 1 does not limit the scope of the present disclosure. Whether done through minimization or maximization, the optimization wrapper (308) identifies the set (or sets) of well operational parameters and the set (or sets) or lithium extraction configuration parameters that optimize hydrocarbon production and lithium extraction according to the trained AI model (e.g., trained AI model A (305)).

An oil and gas field may be subject to constraints, such as safety limits imposed on various devices and sub-processes of an oil and gas field. For example, it may be determined that in order for an oil and gas field to operate safely, pressure, as measured by a given field device, should not exceed a prescribed value. In FIG. 3A, the constraints are referenced as device constraints. In one or more embodiments, the field data (or at least the well data) is monitored by a control system (e.g., SCADA system (126)) such that device and process limits are monitored and controlled by the control system. The optimization wrapper (308) cannot elect any set of well operational parameters and set of lithium extraction configuration parameters that cause any portion of the oil and gas field to exceed pre-defined device constraints. Additional examples of constraints applied to the optimization may include predefined maximum allowable injection rates, production volume ratios, and the maximum quantity of lithium able to be processed. Embodiments of the instant disclosure further allow for distributed decision making. For example, decisions regarding the processing and the injection of water can be made independent of each other, as the produced water and lithium can be re-injected or otherwise utilized. In one or more embodiments, decision-based constraints and distributed decision strategies are accounted for during optimization (e.g., through constraints of the decoupling (i.e., independence) of various operation parameters and lithium extraction configuration parameters.)

FIG. 3B depicts another embodiment of the present invention, where the AI model is used more directly to determine the set of well operational parameters and the set of lithium extraction configuration parameters that jointly optimize hydrocarbon production and lithium extraction in view of an optimization weighting factor (which may be an array). In this embodiment, field data (e.g., Field Data R (310)) is passed to the AI model (e.g., AI Model B (377)). The AI model performs various operations, encompassed by Block 312, described as follows. The AI model (e.g., AI Model B (377)) selects a set of well operational parameters and a set of lithium extraction configuration parameters denoted as selected parameter sets (314) in FIG. 3B. Herein, selection indicates that a value or option is specified for each of the parameters in the set of well operational parameters and the set of lithium extraction configuration parameters. It is noted that the selected parameters sets (314) must obey any constraints (e.g., Device Constrains S (320)) associated with field devices of the oil and gas field. In the embodiment depicted in FIG. 3B, the AI model (e.g., AI Model B (377)) uses the field data (e.g., Field Data R (310)) and checks, as depicted in Block 314, if the oil and gas field is operating at optimal conditions with respect to both the hydrocarbon production and lithium extraction. As previously described with respect to FIG. 3A, joint optimization of the hydrocarbon production and lithium extraction may be performed using an optimization weighting factor. If the hydrocarbon production and lithium extraction are determined to be optimal, the selected parameter sets (314) are retained in Block 316. If, however, the hydrocarbon production and lithium extraction are sub-optimal, the AI model (e.g., AI Model B (377)) intelligently elects selected parameter sets (314) in Block 318. The new selected parameter sets (314) may be intelligently elected through any known method in the art, such as a grid search to probe various parameters in both or either the set of well operational parameters and the set of lithium extraction configuration parameters in search for the optimal selected parameter sets.

Other intelligent search methods, or the AI model (e.g., AI Model B (377)), may include a genetic algorithm, Bayesian search, or a Gaussian process. For example, while a full description of a Gaussian process exceeds the scope of this disclosure, it may simply be said that a Gaussian process is an artificial intelligence method, which in the present case may be used to construct a relationship between hydrocarbon production, lithium extraction and the set of well operational parameters and the set of lithium extraction configuration parameters given the field data. Such a relationship may be mathematically described as

y → = f ⁡ ( x → ❘ D ) ∼ N ( ) , ( 2 ) ,

where {right arrow over (y)} is a vector of quantities representative of, at least, hydrocarbon production and lithium extraction, {right arrow over (x)} is a vector including all the parameters of the set of well operational parameters and the set of lithium extraction configuration parameters, and D is the field data. The output {right arrow over (y)} of a Gaussian process for a given input {right arrow over (x)} will follow a normal distribution with a mean value and a variance. Because the outputs of a Gaussian process follow a normal distribution, the Gaussian process naturally lends itself to uncertainty quantification. As such, the domain of inputs {right arrow over (x)} may be intelligently searched to discover the optimal outputs {right arrow over (y)} within the bounds of uncertainty.

Once elected, the new selected parameters sets (314) are used in the oil and gas field. This process is repeated until the optimal settings have been discovered. Again, like the embodiment of FIG. 3A, the field data (e.g., Field Data R (310)) is continuously monitored by the control system of the oil and gas field (e.g., SCADA system (126)), and all parameters of the selected parameters sets (314) are subjected to any pre-defined system constraints (e.g., Device Constraints S (320)).

In accordance with one or more embodiments, the procedures depicted in FIGS. 3A and 3B may be combined and/or used in complimentary fashion. For example, the search method of the AI model of FIG. 3B (e.g., AI Model B (377)) can be used as the optimization wrapper (308) of the embodiment shown in FIG. 3A. In another embodiment, the procedure of FIG. 3B is used to efficiently probe a domain of inputs spanned by all possible selected parameter sets and record the associated field data, including measurements of hydrocarbon production and lithium extraction, to generate robust training data for the AI model of FIG. 3A (e.g., AI Model A (305)).

The process of evaluating field data and determining the set of well operational parameters and the set of lithium extraction configuration parameters that simultaneously optimize hydrocarbon production and lithium extraction, at least in view of an optimization weighting factor, of an oil and gas field is summarized in the flow chart of FIG. 4. In Block 402, field data for an oil and gas field is obtained. The field data may include measurements from a plurality of devices disposed throughout the oil and gas field. The devices may include various valves which control the flow of fluid throughout the oil and gas field and sensors which measure and quantify the state of various components of the system. The field data may further include a number of wells in the oil and gas field as well as spatial data indicating the location of each well. The spatial data may further include an indication of the distance between each well and, additionally, wellbore paths and wellbore geometry for each well. The field data may further include petrophysical information for the oil and gas field such as one or more subsurface models that digitally represent the distribution of a property (e.g., porosity) throughout the subsurface. The field data may further include well history data for each well including hydrocarbon production and lithium extraction quantities and/or rates as detected and/or measured by one or more field devices. The field day may further include water quality data indicative of, at least, a concentration of lithium in produced water emanating from at least one well of the oil and gas field.

In one or more embodiments, the field data is pre-processed. Pre-processing may include numericalizing the data, scaling the data, selecting features from the data, and engineering features from the data.

In Block 404, a set of well operational parameters is obtained. The set of well operational parameters includes all well control parameters for each well of the oil and gas field. For example, the set of well operational parameters may can include water injection rates, water injection salinity, producer choke rate, etc.

In Block 406, a set of lithium extraction configuration parameters is obtained. The set of lithium extraction configuration parameters includes all configuration parameters of every lithium extraction system of the oil and gas field.

In Block 408, the field data, set of well operational parameters, and set of lithium extraction configuration parameters are processed by an artificial intelligence (AI) model to predict hydrocarbon production and lithium extraction for the oil and gas field. The AI model can also accept, or be informed by, additional information, such as the reservoir formation geology and expected lithium concentration. Various embodiments of the AI model have been described with regards to FIGS. 3A and 3B. The AI model outputs a set of well operational parameters and a set of lithium extraction configuration parameters that, if implemented, according to the AI model will optimize hydrocarbon production and lithium extraction in the oil and gas field.

In Block 410, the set of well operational parameters and the set of lithium extraction configuration parameters are adjusted (e.g., through a control system of the oil and gas field) to their optimal values as determined using the AI model. This adjustment may be performed automatically and autonomously, or may be done manually, or may be checked by a “human-in-the-loop.” For example, adjustments can be made, automatically, to water injection rates and producer choke rate(s) for the various injector and producer wells of an oil and gas field as determined by the AI model and optimization.

Once the set of well operational parameters and the set of lithium extraction configuration parameters have been adjusted, in one or more embodiments, the hydrocarbon production and lithium extraction are monitored by at least one field device from the plurality of oil and gas field devices, as depicted in Block 412. By monitoring the hydrocarbon production and lithium extraction before and after the adjustment of the set of well operational parameters and the set of lithium extraction configuration parameters, the effect of the adjustment on the hydrocarbon production and lithium extraction may be quantified. As such, as shown in Block 412, the set of well operational parameters and the set of lithium extraction configuration parameters may be validated. If the adjusted set of well operational parameters and set of lithium extraction configuration parameters are not found to improve the hydrocarbon production and lithium extraction, the original, or previous, sets of well operational parameters and lithium extraction configuration parameters may be restored. In this case, a AI model may be selected, or the AI model may be re-trained with additional field data.

FIGS. 5A and 5B each depict a flowchart outlining the use of an artificial intelligence-based model to optimize hydrocarbon production and lithium extraction from an oil and gas field, in accordance with one or more embodiments. Specifically, FIG. 5A is directed to oil and gas fields accessing an unconventional reservoir and FIG. 5B is directed toward oil and gas fields accessing a conventional reservoir. Many of the steps are similar between FIGS. 5A and 5B. However, unique numeric labels are used between FIGS. 5A and 5B to emphasize that each figure represents a unique framework (i.e., for an unconventional reservoir and for a conventional reservoir, respectively). Further, while the steps may be similar it should be understood that data and model objects are different and unique between the frameworks. In other words, the oil and gas field referenced in FIG. 5A is likely not the same oil and gas field referenced in FIG. 5B.

As stated, FIG. 5A depicts a flowchart outlining the use of an artificial intelligence (AI)-based model to optimize hydrocarbon production and lithium extraction from an oil and gas field accessing an unconventional reservoir, in accordance with one or more embodiments. In Block 502, field data for the oil and gas field is collected. Field data can further include the number of wells, well data history including production and injection history, water quality data (e.g., lithium concentration in produced water and other geochemical analyses), and well distances, amongst other information and data of the oil and gas field (e.g., petrophysical information, subsurface models, etc.). Additionally, well control parameters (e.g., set of well operational parameters) and lithium extraction configuration parameters (e.g., set of lithium extraction configuration parameters) are identified. In accordance with one or more embodiments, the lithium extraction configuration parameters include, at least, a lithium membrane quality and status. That is, in one or more embodiments, the lithium extraction configuration parameters include an identification of a lithium membrane (e.g., material) as well the use history or status of the lithium membrane (e.g., number of hours in operation). In one or more embodiments, the lithium extraction configuration parameters further include, for example, when using a membrane-based lithium extraction system, the pressure applied across the membrane, selection of membrane pore size (e.g., nanofiltration), electrical field strength (selective electrodialysis), and directing the application and quantity of heat (membrane distillation crystallization).

In Block 504, the field data is filtered for outliers. In one or more embodiments, the field data are filtered based on an z-outlier filtering approach incorporating the removal of erroneous and missing data. In Block 505, an AI model is developed (e.g., trained) to forecast oil and gas rates as well as lithium recovery from the expected water production, water quality, and lithium membrane recovery (i.e., field data, lithium extraction configuration parameters, and well control parameters).

In Block 506, a feature impact analysis based on the calculation of Shapley parameters is utilized in order to determine whether the AI model exhibits signs of overfitting and to determine the main input features. In Block 510, the AI model is integrated into a hydraulic fracture optimization framework such as the optimization wrapper (308) of FIG. 3A, or an existing hydraulic fraction optimization framework such as a reservoir simulator, for the maximization of both oil and gas rates as well as lithium recovery subject to well constraints. In this instance, hydraulic fracturing operation parameters, re-fracturing time periods, and a knowledge of produced water constituents and their concentrations, are included into the estimation of the rates (hydrocarbon and lithium).

Block 512 represents a decision where it is determined if the parameters evaluated in Block 510 are optimal (i.e., at least residing on a Pareto front or efficient frontier, if not a particularly defined point on the Pareto front according to an optimization weighting factor). If the parameters are not determined to be optimal, the optimization step of Block 510 is performed again. Otherwise, the flowchart of FIG. 5A proceeds to Block 514.

In Block 514, the optimal set of parameters (e.g., well control parameters and lithium extraction configuration parameters) identified in Blocks 510 and 512 are applied to the oil and gas field accessing an unconventional reservoir. That is, the optimal parameters define an optimized unconventional reservoir strategy and this unconventional reservoir strategy is incorporated into the operation of the oil and gas field accessing at least one unconventional reservoir. Finally, in Block 516, the AI model is re-trained according to some pre-defined periodicity or indication of AI model performance degradation. For example, in one or more embodiments, the AI model is re-trained with newly acquired field data including hydrocarbon production and lithium extraction data every X months, where X is provided by a user or subject matter expert. Further, in Block 516, the flowchart reverts back to the optimization step of Block 510 to re-optimize oil and gas field operations according to some pre-defined periodicity. In one or more embodiments, re-optimization occurs every Y months, where Y is provided by a user or subject matter expert. Note that periodicity of re-training and re-optimization (i.e., the values of X and Y) need not be the same.

As stated, FIG. 5B depicts a flowchart outlining the use of an artificial intelligence (AI)-based model to optimize hydrocarbon production and lithium extraction from an oil and gas field accessing a conventional reservoir, in accordance with one or more embodiments. In Block 518, field data for the oil and gas field is collected. Field data can further include the number of wells, well data history including production and injection history, water quality data (e.g., lithium concentration in produced water and other geochemical analyses), and well distances, amongst other information and data of the oil and gas field (e.g., petrophysical information, subsurface models, etc.). Additionally, well control parameters (e.g., set of well operational parameters) and lithium extraction configuration parameters (e.g., set of lithium extraction configuration parameters) are identified. In accordance with one or more embodiments, the lithium extraction configuration parameters include, at least, a lithium membrane quality and status. That is, in one or more embodiments, the lithium extraction configuration parameters include an identification of a lithium membrane (e.g., material) as well the use history or status of the lithium membrane (e.g., number of hours in operation).

In Block 520, the field data is filtered for outliers. In one or more embodiments, the field data are filtered based on an z-outlier filtering approach incorporating the removal of erroneous and missing data. In Block 521, an AI model is developed (e.g., trained) to forecast oil and gas rates as well as lithium recovery from the expected water production, water quality, and lithium membrane recovery (i.e., field data, lithium extraction configuration parameters, and well control parameters).

In Block 522, a feature impact analysis based on the calculation of Shapley parameters is utilized in order to determine whether the AI model exhibits signs of overfitting and to determine the main input features. In Block 524, the AI model is integrated into a well injection optimization framework for the maximization of both oil and gas rates as well as lithium recovery subject to well constraints. The well injection optimization framework may be the optimization wrapper (308) of FIG. 3A or an existing well injection optimization framework such as a reservoir simulator, for the maximization of both oil and gas rates as well as lithium recovery subject to well constraints. In this instance, well injection parameters are included into the estimation of the rates (hydrocarbon and lithium).

Block 526 represents a decision where it is determined if the parameters evaluated in Block 524 are optimal (i.e., at least residing on a Pareto front or efficient frontier, if not residing at a particular point on the Pareto front according to an optimization weighting factor). If the parameters are not determined to be optimal, the optimization step of Block 524 is performed again. Otherwise, the flowchart of FIG. 5B proceeds to Block 528.

In Block 528, the optimal set of parameters (e.g., well control parameters and lithium extraction configuration parameters) identified in Blocks 524 and 526 are applied to the oil and gas field accessing a conventional reservoir. That is, the optimal parameters define an optimized conventional reservoir strategy and this conventional reservoir strategy is incorporated into the operation of the oil and gas field accessing at least one conventional reservoir. Finally, in Block 530, the AI model is re-trained according to some pre-defined periodicity or indication of AI model performance degradation. For example, in one or more embodiments, the AI model is re-trained with newly acquired field data including hydrocarbon production and lithium extraction data every U months, where U is provided by a user or subject matter expert. Further, in Block 530, the flowchart reverts back to the optimization step of Block 524 to re-optimize oil and gas field operations according to some pre-defined periodicity. In one or more embodiments, re-optimization occurs every V months, where V is provided by a user or subject matter expert. Note that periodicity of re-training and re-optimization (i.e., the values of U and V) need not be the same.

While the various blocks in FIGS. 4, 5A, and 5B are presented and described sequentially, one of ordinary skill in the art will appreciate that some or all of the blocks may be executed in different orders, may be combined or omitted, and some or all of the blocks may be executed in parallel. Furthermore, the blocks may be performed actively or passively.

Embodiments of the present disclosure may provide at least one of the following advantages. As noted, complex interactions between oil and gas field components and sub-processes exist such that configuring a plurality of devices, including devices, settings, and/or the configuration of one or more lithium extraction system, for simultaneously (or jointly) optimal hydrocarbon production and lithium extraction is a difficult task. Further, the state and behavior of oil and gas fields is transient over the lifetime of the constituent wells requiring continual changes to the plurality of field devices to enhance hydrocarbon production and lithium extraction. By continuously receiving and processing field data with an AI model, the oil and gas field can be maintained in an optimal state greatly reducing the cost and time required to identify optimal settings which change with the transient nature of the wells. This, in turn, improves hydrocarbon (i.e., oil and gas) and lithium yields and prolongs the life of constituent wells.

In accordance with one or more embodiments, one or more of the AI models (206) discussed herein, such as AI Model A (305) is a neural network. A diagram of a neural network is shown in FIG. 6. At a high level, a neural network (600) may be graphically depicted as being composed of nodes (602), where here any circle represents a node, and edges (604), shown here as directed lines. The nodes (602) may be grouped to form layers (605). FIG. 6 displays four layers (608, 610, 612, 614) of nodes (602) where the nodes (602) are grouped into columns, however, the grouping need not be as shown in FIG. 6. The edges (604) connect the nodes (602). Edges (604) may connect, or not connect, to any node(s) (602) regardless of which layer (605) the node(s) (602) is in. That is, the nodes (602) may be sparsely and residually connected. A neural network (600) will have at least two layers (605), where the first layer (608) is considered the “input layer” and the last layer (614) is the “output layer.” Any intermediate layer (610, 612) is usually described as a “hidden layer”. A neural network (600) may have zero or more hidden layers (610, 612) and a neural network (600) with at least one hidden layer (610, 612) may be described as a “deep” neural network or as a “deep learning method.” In general, a neural network (600) may have more than one node (602) in the output layer (614). In this case the neural network (600) may be referred to as a “multi-target” or “multi-output” network.

Nodes (602) and edges (604) carry additional associations. Namely, every edge is associated with a numerical value. The edge numerical values, or even the edges (604) themselves, are often referred to as “weights” or “parameters.” While training a neural network (600), numerical values are assigned to each edge (604). Additionally, every node (602) is associated with a numerical variable and an activation function. Activation functions are not limited to any functional class, but traditionally follow the form

A = f ⁡ ( ∑ i ∈ ( incoming ) [ ( node ⁢ value ) i ⁢ ( edge ⁢ value ) i ] ) , ( 3 ) ,

where i is an index that spans the set of “incoming” nodes (602) and edges (604) and f is a user-defined function. Incoming nodes (602) are those that, when viewed as a graph (as in FIG. 6), have directed arrows that point to the node (602) where the numerical value is being computed. Some functions for ƒ may include the linear function ƒ(x)=x, sigmoid function

f ⁡ ( x ) = 1 1 + e - x ,

and rectified linear unit function ƒ(x)=max(0, x), however, many additional functions are commonly employed. Every node (602) in a neural network (600) may have a different associated activation function. Often, as a shorthand, activation functions are described by the function ƒ by which it is composed. That is, an activation function composed of a linear function ƒ may simply be referred to as a linear activation function without undue ambiguity.

When the neural network (600) receives an input, the input is propagated through the network according to the activation functions and incoming node (602) values and edge (604) values to compute a value for each node (602). That is, the numerical value for each node (602) may change for each received input. Occasionally, nodes (602) are assigned fixed numerical values, such as the value of 1, that are not affected by the input or altered according to edge (604) values and activation functions. Fixed nodes (602) are often referred to as “biases” or “bias nodes” (606), displayed in FIG. 6 with a dashed circle.

In some implementations, the neural network (600) may contain specialized layers (605), such as a normalization layer, or additional connection procedures, like concatenation. One skilled in the art will appreciate that these alterations do not exceed the scope of this disclosure.

As noted, the training procedure for the neural network (600) comprises assigning values to the edges (604). To begin training the edges (604) are assigned initial values. These values may be assigned randomly, assigned according to a prescribed distribution, assigned manually, or by some other assignment mechanism. Once edge (604) values have been initialized, the neural network (600) may act as a function, such that it may receive inputs and produce an output. As such, at least one input is propagated through the neural network (600) to produce an output. Recall, that a given data set will be composed of inputs and associated target(s), where the target(s) represent the “ground truth,” or the otherwise desired output. In accordance with one or more embodiments, the input of the neural network is the field data (which may be pre-processed), set of well operational parameters, and set of lithium extraction configuration parameters and the targets are hydrocarbon production and lithium extraction (given as either quantities over a pre-defined period or rates).

The neural network (600) output is compared to the associated input data target(s). The comparison of the neural network (600) output to the target(s) is typically performed by a so-called “loss function;” although other names for this comparison function such as “error function,” “misfit function,” and “cost function” are commonly employed. Many types of loss functions are available, such as the mean-squared-error function, however, the general characteristic of a loss function is that the loss function provides a numerical evaluation of the similarity between the neural network (600) output and the associated target(s). The loss function may also be constructed to impose additional constraints on the values assumed by the edges (604), for example, by adding a penalty term, which may be physics-based, or a regularization term. Generally, the goal of a training procedure is to alter the edge (604) values to promote similarity between the neural network (600) output and associated target(s) over the data set. Thus, the loss function is used to guide changes made to the edge (604) values, typically through a process called “backpropagation.”

While a full review of the backpropagation process exceeds the scope of this disclosure, a brief summary is provided. Backpropagation consists of computing the gradient of the loss function over the edge (604) values. The gradient indicates the direction of change in the edge (604) values that results in the greatest change to the loss function. Because the gradient is local to the current edge (604) values, the edge (604) values are typically updated by a “step” in the direction indicated by the gradient. The step size is often referred to as the “learning rate” and need not remain fixed during the training process. Additionally, the step size and direction may be informed by previously seen edge (604) values or previously computed gradients. Such methods for determining the step direction are usually referred to as “momentum” based methods.

Once the edge (604) values have been updated, or altered from their initial values, through a backpropagation step, the neural network (600) will likely produce different outputs. Thus, the procedure of propagating at least one input through the neural network (600), comparing the neural network (600) output with the associated target(s) with a loss function, computing the gradient of the loss function with respect to the edge (604) values, and updating the edge (604) values with a step guided by the gradient, is repeated until a termination criterion is reached. Common termination criteria are: reaching a fixed number of edge (604) updates, otherwise known as an iteration counter; a diminishing learning rate; noting no appreciable change in the loss function between iterations; reaching a specified performance metric as evaluated on the data or a separate hold-out data set. Once the termination criterion is satisfied, and the edge (604) values are no longer intended to be altered, the neural network (600) is said to be “trained”.

While multiple embodiments using different AI models have been suggested, one skilled in the art will appreciate that this process, of determining the set of well operational parameters and set of lithium extraction configuration parameters that optimize both hydrocarbon production and lithium extraction, is not limited to the listed AI models. AI models such as a random forest, support vector machines, or non-parametric methods such as K-nearest neighbors may be readily inserted into this framework and do not depart from the scope of this disclosure.

In accordance with one or more embodiments, the AI model (e.g., AI Model A (305)) used in the frameworks described herein is a long short-term memory (LSTM) network. To best understand a LSTM network, it is helpful to describe the more general recurrent neural network, for which an LSTM may be considered a specific implementation.

FIG. 7A depicts the general structure of a recurrent neural network (RNN). An RNN is graphically composed of an RNN Block (710) and a recurrent connection (750). The RNN Block may be thought of as a function which accepts an Input (720) and a State (730) and produces an Output (740). Without loss of generality, such a function may be written as

Output=RNN Block(Input,State).  (4)

The RNN Block (710) generally comprises one or more matrices and one or more bias vectors. The elements of the matrices and bias vectors are commonly referred to as “weights” or “parameters” in the literature such that the matrices may be referenced as weight matrices or parameter matrices without ambiguity. It is noted that for situations with higher dimensional inputs (e.g. inputs with a tensor rank greater than or equal to 2), the weights of an RNN Block (710) may be contained in higher order tensors, rather than in matrices or vectors. For clarity, the present example will consider Inputs (720) as vectors or as scalars such that the RNN Block (710) comprises one or more weight matrices and bias vectors, however, one with ordinary skill in the art will appreciate that this choice does not impose a limitation on the present disclosure. Typically, an RNN Block (710) has two weight matrices and a single bias vector which are distinguished with an arbitrary naming nomenclature. A commonly employed naming convention is to call one weight matrix W and the other U and to reference the bias vector as b.

An important aspect of an RNN is that it is intended to process sequential, or ordered, data; for example, a time-series. In the RNN, the Input (720) may be considered a single part of a sequence. As an illustration, consider a sequence composed of Y parts. Each part may be considered an input, indexed by t, such that the sequence may be written as sequence=[input₁, input₂, input_t, . . . , input_Y-1, input_Y]. Each Input (720) (e.g., input₁ of a sequence) may be a scalar, vector, matrix, or higher-order tensor. Recall that a given seismic data set is composed of N_c traces (or channels) and N_t discrete time steps. In accordance with one or more embodiments, each Input (720) (or element of a sequence) is an array of traces at a single time step. That, each Input (1020) is considered a vector with N_c elements.

To process a sequence, an RNN receives the first ordered Input (720) of the sequence, input₁, along with a State (730), and processes them with the RNN Block (710) according to EQ. 4 to produce an Output (740). The Output (740) may be a scalar, vector, matrix, or tensor of any rank. For the present example, the Output (1040) is considered a vector with k elements. The State (730) is of the same type and size as the Output (740) (e.g., a vector with k elements). For the first ordered input, the State (730) is usually initialized with all of its elements set to the value zero. For the second ordered Input (720), input₂, of the sequence, the Input (720) is processed similarly according to EQ. 4, however, the State (730) received by the RNN Block (710) is set to the value of the Output (740) determined when processing the first ordered Input (720). This process of assigning the State (730) the value of the last produced Output (740) is depicted with the recurrent connection (750) in FIG. 7A. All the Inputs (720) in a sequence are processed by the RNN Block (710) in this manner; that is, the State (730) associated with an Input (720) is the Output (740) of the RNN Block (710) produced by the previous Input (720) (with the exception of the first Input (720) in the sequence). In some implementations, each Output (740), one for each Input (710) within a sequence, is stored for later processing and use. In other implementations, only the final Output (740), or the Output (740) which is produced when the Input (720) input_Y is processed by the RNN Block (710), is retained.

In greater detail, the process of the RNN Block (710), or EQ. 4, may be generally written as

Output = RNN ⁢ ⁢ Block ( input , state ) = f ⁡ ( U · state + W · input + b → ) , ( 5 )

where W, U, and {right arrow over (b)} are the weight matrices and bias vector of the RNN Block (710), respectively, and f is an “activation function.” Some functions for ƒ may include the sigmoid function

f ⁡ ( x ) = 1 1 + e - x ,

and rectified linear unit (ReLU) function ƒ(x)=max(0, x), however, many additional functions are commonly employed.

To further illustrate a RNN, a pseudo-code implementation of a RNN is as follows.

RNN Algorithm

	Note:
	Nc = input length
	k = output length
	W ∈   kxk
	∪ ∈   kxNc
	{right arrow over (b)} ∈   k
	1: state = [01, 02, ..., 0k−1, 0k]T
	2: for input in sequence:
	3:  {right arrow over (z)}1 = matmul(U, state)
	4:  {right arrow over (z)}2 = matmul(W, input)
	5:  output = f ({right arrow over (z)}1 + {right arrow over (z)}2 + {right arrow over (b)})
	6:  state = output

In keeping with the previous examples, both the inputs and the outputs are considered vectors of lengths N_c and k, respectively, however, in general, this need not be the case. With the lengths of these vectors defined, the shapes of the weight matrices, bias vector, and State (730) vector may be specified. To begin processing a sequence, the State (730) vector is initialized with values of zero as shown in line 1 of the pseudo-code. Note that in some implementations, the number of inputs contained within a sequence may not be known or may vary between sequences. One with ordinary skill in the art will recognize that an RNN may be implemented without knowing, beforehand, the length of the sequence to be processed. This is demonstrated in line 2 of the pseudo-code by indicating that each input in the sequence will be processed sequentially without specifying the number of inputs in the sequence. Once an Input (720) is received, a matrix multiplication operator is applied between the weight matrix U and the State (730) vector. The resulting product is assigned to the temporary variable {right arrow over (z)}₁. Likewise, a matrix multiplication operator is applied between the weight matrix W and the Input (710) with the result assigned to the variable {right arrow over (z)}₂. For the present example, due the Input (720) and Output (740) each being defined as vectors, the products in lines 3 and 4 of the pseudo-code may be expressed as matrix multiplications, however, in general, the dot product between the weight matrix and corresponding State (730) or Input (720) may be applied. The Output (740) is determined by summing {right arrow over (z)}₁, {right arrow over (z)}₂, and the bias vector b and applying the activation function ƒ elementwise. The State (730) is set to the Output (740) and the whole process is repeated until each Input (720) in a sequence has been processed.

FIG. 7B depicts an “unrolled” version of the RNN of FIG. 7A. Unrolling the RNN allows one to see how the sequential inputs, indexed by t, produce sequential outputs and how the state is passed through various inputs of the sequence. It is noted that while the “unrolled” depiction shows multiple RNN Blocks (710), these blocks are the same such that they are comprised of the same weight matrices and bias vector.

As previously stated, generally, training a machine-learned model requires that pairs of inputs and one or more targets (i.e., a training dataset) are passed to the machine-learned model. During this process the machine-learned model “learns” a representative model which maps the received inputs to the associated outputs. In the context of an RNN, the RNN receives a sequence, wherein the sequence can be partitioned into one or more sequential parts (Inputs (720) above), and maps the sequence to an overall output, which may also be a sequence. To remove ambiguity and distinguish the overall output of an RNN from any intermediate Outputs (740) produced by the RNN Block (710), the overall output will be referred to herein as a RNN result. In other words, an RNN receives a sequence and returns a RNN result. The training procedure for a RNN comprises assigning values to the weight matrices and bias vector of the RNN Block (710). For brevity, the elements of the weight matrices and bias vector will be collectively referred to as the RNN weights. To begin training the RNN weights are assigned initial values. These values may be assigned randomly, assigned according to a prescribed distribution, assigned manually, or by some other assignment mechanism. Once the RNN weights have been initialized, the RNN may act as a function, such that it may receive a sequence and produce a RNN result. As such, at least one sequence may be propagated through the RNN to produce a RNN result. For training, a training dataset is composed of one or more sequences and desired RNN results, where the desired RNN results represent the “ground truth”, or the true RNN results that should be returned for the given sequences. For clarity, and consistency with previous discussions of machine-learned model training, the desired or true RNN results will be referred to as targets. When processing sequences, the RNN result produced by the RNN is compared to the associated target. The comparison of a RNN result to the target(s) is typically performed by a loss function. As before, other names for this comparison function such as “error function” and “cost function” are commonly employed. Many types of loss functions are available, such as the mean squared error function, however, the general characteristic of a loss function is that the loss function provides a numerical evaluation of the similarity between the RNN result and the associated target(s). The loss function may also be constructed to impose additional constraints on the values assumed by RNN weights, for example, by adding a penalty term, which may be physics-based, or a regularization term. Generally, the goal of a training procedure is to alter the RNN weights to promote similarity between the RNN results and associated targets over the training dataset. Thus, the loss function is used to guide changes made to the RNN weights, typically through a process called “backpropagation through time.”

A long short-term memory (LSTM) network may be considered a specific, and more complex, instance of a recurrent neural network (RNN). FIG. 7C is an unrolled depiction of a LSTM where the internal components of the LSTM are displayed as labelled abstractions. A LSTM, like a RNN, has a recurrent connection, such that the output produced by a single input in a sequence is forwarded as the state to be used with the subsequent input. However, an LSTM also possesses another “state-like” data structure commonly referred to as the “carry.” The carry, like the state and input may be a scalar, vector, matrix, or tensor of any rank depending on the context of the application. Like unto the description of the RNN, for simplicity, the carry will be considered a vector in the following discussion of the LSTM. The LSTM receives an input, state, and carry and produces an output and a new carry. The output and the new carry are passed to the LSTM as the state and carry for the subsequent input. This sequential process, indexed by t, may be described functionally as

( output t , carry t ) = LSTM ⁢ Block ( input t , carry t - 1 , state t ) = LSTM ⁢ Block ( input t , carry t - 1 , output t - 1 ) , ( 6 )

where the LSTM Block, like the RNN Block, comprises one or more weight matrices and bias vectors and the processing steps necessary to transform an input, state, and carry to an output and new carry.

LSTMs may be configured in a variety of ways, however, the processes depicted in FIG. 7C are the most common. As shown in FIG. 7C, an LSTM Block receives an input (input_t), a state (state_t), and a carry (carry_t-1). Again, assuming that the inputs, carry, and outputs are all vectors, the weights of the LSTM Block may be considered to reside in eight matrices and four bias vectors. These matrices and vectors are conventionally named W_i, U_i, W_f, U_f, W_c, U_c, W_o, U_o and {right arrow over (b)}_i, {right arrow over (b)}_f, {right arrow over (b)}_c, {right arrow over (b)}_o, respectively. The processes of the LSTM Block are as follows. Block 760 represents the following first operation

f → = a 1 ( U f · state t + W f · input c + b → f ) , ( 7 )

• • where a₁ is an activation function applied elementwise to the result of the parenthetical expression and the resulting vector is {right arrow over (ƒ)}. Block 765 implements the following second operation

ι → = a 3 ( U i · state t + W t · input t + b → i ) , ( 8 )

• • where a₂ is an activation function which may be the same or different to @1 and is applied elementwise to the result of the parenthetical expression. The resulting vector is {right arrow over (t)}. Block 770 implements the following third operation

c → = a 3 ( U c · state t + W c · input t + b → c ) , ( 9 )

• • where a₃ is an activation function which may be the same or different to either a₁ or a₂ and is applied elementwise to the result of the parenthetical expression. The resulting vector is {right arrow over (c)}. In block 775, vectors {right arrow over (i)} and {right arrow over (c)} are multiplied according to a fourth operation

z → 3 = ι → ⊙ c → , ( 10 )

• • where ⊙ indicates the Hadamard product (i.e., elementwise multiplication). Likewise, in block 785 the carry vector from the previous sequential input (carry_t-1) vector and the vector/are multiplied according to a fifth operation

z → 4 = carry t - 1 ⊙ f → . ( 11 )

• • The results of the operations of blocks 775 and 785 ({right arrow over (z)}₃ and {right arrow over (z)}₄, respectively) are added together in block 780, a sixth operation, to form the new carry (carry_t)

carry t = z → 3 + z → 4 . ( 12 )

• • In block 790, the current input and state vectors are processed according to a seventh operation

o → = a 4 ( U o · state t + W o · input t + b → o ) , ( 13 )

• • where a₄ is an activation function which may be unique or identical to any other used activation function and is applied elementwise to the result of the parenthetical expression. The result is the vector {right arrow over (o)}. In block 795, an eighth operation, the new carry (carry_t) is passed through an activation function a₅. The activation a₅ is usually the hyperbolic tangent function but may be any known activation function. The eighth operations (block 795) may be represented as

z → 5 = a 5 ( carry t ) . ( 14 )

• • Finally, the output of the LSTM Block (output) is determined in block 798 by taking the Hadamard product of {right arrow over (z)}₅ and {right arrow over (o)}, a ninth operation shown mathematically as

output t = z → 5 ⊙ o → . ( 15 )

• • The output of the LSTM Block is used as the state vector for the subsequent input. Again, as in the case of the RNN, the outputs of the LSTM Block applied to a sequence of inputs may be stored and further processed or, in some implementations, only the final output is retained. While the processes of the LSTM Block described above used vector inputs and outputs, it is emphasized that an LSTM network may be applied to sequences of any dimensionality. In these circumstances the rank and size of the weight tensors will change accordingly. One with ordinary skill in the art will recognize that there are many alterations and variations that can be made to the general LSTM structure described herein, such that the description provided does not impose a limitation on the present disclosure.

Embodiments may be implemented on a computer system. FIG. 8 is a block diagram of a computer system (802) used to provide computational functionalities associated with described algorithms, methods, functions, processes, flows, and procedures as described in the instant disclosure, according to one or more embodiments. The illustrated computer (802) is intended to encompass any computing device such as a server, desktop computer, laptop/notebook computer, wireless data port, smart phone, personal data assistant (PDA), tablet computing device, one or more processors within these devices, or any other suitable processing device such as an edge computing device, including both physical or virtual instances (or both) of the computing device. An edge computing device is a dedicated computing device that is, typically, physically adjacent to the process or control with which it interacts. For example, the AI model may be implemented on an edge computing device in order to quickly provide optimal sets of well operational parameters and lithium extraction configuration parameters to associated field devices or their controllers (e.g., control system).

Additionally, the computer (802) may include a computer that includes an input device, such as a keypad, keyboard, touch screen, or other device that can accept user information, and an output device that conveys information associated with the operation of the computer (802), including digital data, visual, or audio information (or a combination of information), or a GUI.

The computer (802) can serve in a role as a client, network component, a server, a database or other persistency, or any other component (or a combination of roles) of a computer system for performing the subject matter described in the instant disclosure. In some implementations, one or more components of the computer (802) may be configured to operate within environments, including cloud-computing-based, local, global, or other environment (or a combination of environments).

At a high level, the computer (802) is an electronic computing device operable to receive, transmit, process, store, or manage data and information associated with the described subject matter. According to some implementations, the computer (802) may also include or be communicably coupled with an application server, e-mail server, web server, caching server, streaming data server, business intelligence (BI) server, or other server (or a combination of servers).

The computer (802) can receive requests over network (830) from a client application (for example, executing on another computer (802) and responding to the received requests by processing the said requests in an appropriate software application. In addition, requests may also be sent to the computer (802) from internal users (for example, from a command console or by other appropriate access method), external or third-parties, other automated applications, as well as any other appropriate entities, individuals, systems, or computers.

Each of the components of the computer (802) can communicate using a system bus (803). In some implementations, any or all of the components of the computer (802), both hardware or software (or a combination of hardware and software), may interface with each other or the interface (804) (or a combination of both) over the system bus (803) using an application programming interface (API) (812) or a service layer (813) (or a combination of the API (812) and service layer (813). The API (812) may include specifications for routines, data structures, and object classes. The API (812) may be either computer-language independent or dependent and refer to a complete interface, a single function, or even a set of APIs. The service layer (813) provides software services to the computer (802) or other components (whether or not illustrated) that are communicably coupled to the computer (802). The functionality of the computer (802) may be accessible for all service consumers using this service layer. Software services, such as those provided by the service layer (813), provide reusable, defined business functionalities through a defined interface. For example, the interface may be software written in JAVA, C++, or other suitable language providing data in extensible markup language (XML) format or another suitable format. While illustrated as an integrated component of the computer (802), alternative implementations may illustrate the API (812) or the service layer (813) as stand-alone components in relation to other components of the computer (802) or other components (whether or not illustrated) that are communicably coupled to the computer (802). Moreover, any or all parts of the API (812) or the service layer (813) may be implemented as child or sub-modules of another software module, enterprise application, or hardware module without departing from the scope of this disclosure.

The computer (802) includes an interface (804). Although illustrated as a single interface (804) in FIG. 8, two or more interfaces (804) may be used according to particular needs, desires, or particular implementations of the computer (802). The interface (804) is used by the computer (802) for communicating with other systems in a distributed environment that are connected to the network (830). Generally, the interface (804) includes logic encoded in software or hardware (or a combination of software and hardware) and operable to communicate with the network (830). More specifically, the interface (804) may include software supporting one or more communication protocols associated with communications such that the network (830) or interface's hardware is operable to communicate physical signals within and outside of the illustrated computer (802).

The computer (802) includes at least one computer processor (805). Although illustrated as a single computer processor (805) in FIG. 8, two or more processors may be used according to particular needs, desires, or particular implementations of the computer (802). Generally, the computer processor (805) executes instructions and manipulates data to perform the operations of the computer (802) and any algorithms, methods, functions, processes, flows, and procedures as described in the instant disclosure.

The computer (802) also includes a memory (806) that holds data for the computer (802) or other components (or a combination of both) that can be connected to the network (830). The memory may be a non-transitory computer readable medium. For example, memory (806) can be a database storing data consistent with this disclosure. Although illustrated as a single memory (806) in FIG. 8, two or more memories may be used according to particular needs, desires, or particular implementations of the computer (802) and the described functionality. While memory (806) is illustrated as an integral component of the computer (802), in alternative implementations, memory (806) can be external to the computer (802).

The application (807) is an algorithmic software engine providing functionality according to particular needs, desires, or particular implementations of the computer (802), particularly with respect to functionality described in this disclosure. For example, application (807) can serve as one or more components, modules, applications, etc. Further, although illustrated as a single application (807), the application (807) may be implemented as multiple applications (807) on the computer (802). In addition, although illustrated as integral to the computer (802), in alternative implementations, the application (807) can be external to the computer (802).

There may be any number of computers (802) associated with, or external to, a computer system containing computer (802), wherein each computer (802) communicates over network (830). Further, the term “client,” “user,” and other appropriate terminology may be used interchangeably as appropriate without departing from the scope of this disclosure. Moreover, this disclosure contemplates that many users may use one computer (802), or that one user may use multiple computers (802).

Although only a few example embodiments have been described in detail above, those skilled in the art will readily appreciate that many modifications are possible in the example embodiments without materially departing from this invention. Accordingly, all such modifications are intended to be included within the scope of this disclosure as defined in the following claims.