METHODS AND SYSTEMS FOR OPTIMIZING GAS-LIFTING OPERATIONS

Abstract

A method for optimal execution of a gas-lifting procedure. The method includes obtaining gas-lift data from a well site with gas-lift, the well site with gas-lift including an oil and gas well with access to a hydrocarbon reservoir and a gas-lift system including a gas pump. The method further includes obtaining a set of gas-lift parameters related to the well site with gas-lift and determining, with a machine learning (ML) model, a predicted gas-lift based on the gas-lift data and the set of gas-lift parameters, and determining, with an optimizer applied to the ML model, an optimal set of gas-lift parameters such that the predicted gas-lift is optimized. The method further includes adjusting, automatically, the set of gas-lift parameters to the optimal set of gas-lift parameters and injecting gas into the well using the gas-lift system according to the optimal set of gas-lift parameters.

Description

BACKGROUND

Hydrocarbon fluids are often found in hydrocarbon reservoirs located in porous rock formations far below the Earth's surface. Wells may be drilled to extract the hydrocarbon fluids from the hydrocarbon reservoirs. Often, hydrocarbon fluids are able to flow naturally through producing wells (i.e., those wells configured for the production of hydrocarbons) because the pressure within the reservoir is high enough to propel the hydrocarbons to the surface. However, when a reservoir becomes depleted, or is naturally a reservoir with low-pressure, gas-lift may be utilized to produce the hydrocarbon fluids. Gas-lift uses a source of high-pressure gas to lower the bulk density of a fluid mixture with hydrocarbons and “lift” the mixture to the surface. In some examples, gas-lift systems use an external source of gas which is injected into production tubing located in the production well. The gas mixes with the hydrocarbon fluids in the production tubing reducing the density of the hydrocarbon fluid and gas mixture until the mixture becomes light enough to flow using the available reservoir pressure. Gas-lifting is challenging, particularly in offshore fields where additional costs are encountered and resources become more difficult to acquire and manage. As such, there exists a need to efficiently operate gas-lifting procedures with high success rates.

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 optimal execution of a gas-lifting procedure. The method includes obtaining gas-lift data from a well site with gas-lift, the well site with gas-lift including an oil and gas well with access to a hydrocarbon reservoir and a gas-lift system including a gas pump. The method further includes obtaining a set of gas-lift parameters related to the well site with gas-lift and determining, with a machine learning (ML) model, a predicted gas-lift based on the gas-lift data and the set of gas-lift parameters, and determining, with an optimizer applied to the ML model, an optimal set of gas-lift parameters such that the predicted gas-lift is optimized. The method further includes adjusting, automatically, the set of gas-lift parameters to the optimal set of gas-lift parameters and injecting gas into the well using the gas-lift system according to the optimal set of gas-lift parameters.

Embodiments disclosed herein generally relate to a system including a well site with gas-lift including an oil and gas well with access to a hydrocarbon reservoir and a gas-lift system including a gas pump. The system further includes a plurality of field devices disposed throughout the well site with gas-lift, the plurality of field devices gathering gas-lift data from the well site with gas-lift, and a control system configured to adjust one or more field devices in the plurality of field devices and the gas-lift system. The system further includes and a computer configured to obtain the gas-lift data from the well and from the reservoir, obtain a set of gas-lift parameters for the well site with gas-lift, determine, with a machine learning (ML) model, a predicted gas-lift based on the gas-lift data and gas-lift parameters, and determine, with an optimizer applied to the ML model, an optimal set of gas-lift parameters such that the predicted gas-lift is optimized. The computer is further configured to adjust, automatically, the set of gas-lift parameters to the optimal set of gas-lift parameters, and transmit a signal to inject gas into the well using the gas-lift system according to the optimal set of gas-lift parameters.

Embodiments disclosed herein generally relate to a non-transitory computer-readable memory comprising computer-executable instructions stored thereon that, when executed on a processor, cause the processor to perform the following steps. The steps include obtaining gas-lift data from a well site with gas-lift, the well site with gas-lift including an oil and gas well with access to a hydrocarbon reservoir and a gas-lift system including a gas pump. The steps further include obtaining a set of gas-lift parameters related to the well site with gas-lift, determining, with a machine learning (ML) model, a predicted gas-lift based on the gas-lift data and the set of gas-lift parameters, and determining, with an optimizer applied to the ML model, an optimal set of gas-lift parameters such that the predicted gas-lift is optimized. The steps further include adjusting, automatically, the set of gas-lift parameters to the optimal set of gas-lift parameters, and transmitting a signal to inject gas into the well using the gas-lift system according to the optimal set of gas-lift parameters.

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

BRIEF DESCRIPTION OF DRAWINGS

Specific embodiments of the disclosed technology will now be described in detail with reference to the accompanying figures. Like elements in the various figures are denoted by like reference numerals for consistency. The sizes and relative positions of elements in the drawings are not necessarily drawn to scale. For example, the shapes of various elements and angles are not necessarily drawn to scale, and some of these elements may be arbitrarily enlarged and positioned to improve drawing legibility. Further, the particular shapes of the elements as drawn are not necessarily intended to convey any information regarding the actual shape of the particular elements and have been solely selected for ease of recognition in the drawing.

FIG. 1 depicts a well site with gas-lift in accordance with one or more embodiments.

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

FIG. 3 depicts an example lift capability plot in accordance with one or more embodiments.

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

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

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

FIG. 7 depicts a system in accordance with one or more embodiments.

DETAILED DESCRIPTION

In the following detailed description of embodiments of the disclosure, numerous specific details are set forth in order to provide a more thorough understanding of the disclosure. However, it will be apparent to one of ordinary skill in the art that the disclosure may be practiced without these specific details. In other instances, well-known features have not been described in detail to avoid unnecessarily complicating the description.

Throughout the application, ordinal numbers (e.g., first, second, third, etc.) may be used as an adjective for an element (i.e., any noun in the application). The use of ordinal numbers is not to imply or create any particular ordering of the elements nor to limit any element to being only a single element unless expressly disclosed, such as using the terms “before,” “after,” “single,” and other such terminology. Rather, the use of ordinal numbers is to distinguish between the elements. By way of an example, a first element is distinct from a second element, and the first element may encompass more than one element and succeed (or precede) the second element in an ordering of elements.

It is to be understood that the singular forms “a,” “an,” and “the” include plural referents unless the context clearly dictates otherwise. For example, an “oil or gas well,” may include any number of “oil or gas wells” without limitation.

Terms such as “approximately,” “substantially,” etc., mean that the recited characteristic, parameter, or value need not be achieved exactly, but that deviations or variations, including for example, tolerances, measurement error, measurement accuracy limitations and other factors known to those of skill in the art, may occur in amounts that do not preclude the effect the characteristic was intended to provide.

It is to be understood that one or more of the steps shown in the flowcharts may be omitted, repeated, and/or performed in a different order than the order shown. Accordingly, the scope disclosed herein should not be considered limited to the specific arrangement of steps shown in the flowcharts.

Although multiple dependent claims are not introduced, it would be apparent to one of ordinary skill that the subject matter of the dependent claims of one or more embodiments may be combined with other dependent claims.

In the following description of FIGS. 1-7, any component described with regard to a figure, in various embodiments disclosed herein, may be equivalent to one or more like-named components described with regard to any other figure. For brevity, descriptions of these components will not be repeated with regard to each figure. Thus, each and every embodiment of the components of each figure is incorporated by reference and assumed to be optionally present within every other figure having one or more like-named components. Additionally, in accordance with various embodiments disclosed herein, any description of the components of a figure is to be interpreted as an optional embodiment which may be implemented in addition to, in conjunction with, or in place of the embodiments described with regard to a corresponding like-named component in any other figure.

In general, embodiments of the disclosure include systems and methods for optimizing gas-lifting operations in oil and gas wells. Gas-lift systems aid or increase hydrocarbon production in hydrocarbon production wells by injecting high-pressure gas, from the surface, down a casing annulus or coil tubing into fluids disposed in the production piping. The fluid density and hydrostatic pressure of the fluid are reduced by the introduction of the injected gas, thereby allowing the in-situ reservoir pressure to lift the lightened fluids. In one or more embodiments, the gas used in a gas-lift system includes an external source of gas and a gas pump. A gas-lift system thus refers to the set of components and mechanical or electrical devices used in a gas-lifting operation.

In accordance with one or more embodiments, systems of the instant disclosure include a well site with gas-lift, including an oil and gas well (“well”) with access to a subsurface formation or reservoir (“reservoir”) and access to an external source of gas to be used for gas-lifting. In one or more embodiments, the external source of gas may be included by a gas-lift system including a gas pump, and the injected gas may be nitrogen. The operation of the external source of gas, with respect to injecting gas into the well, is controlled by the assignation of each parameter of a set of gas-lift parameters to a given state or value. The result of injecting gas into the well is sensitive to a number of factors, including the physical properties or characteristics of the well, the reservoir, and the various fluids that may be present in either the well or reservoir. The result of injecting gas into the well is further sensitive to the values or states of the parameters of the set of gas-lift parameters that define, at least, the operation of the external source of gas and its injection. In one or more embodiments, the set of gas-lift parameters is adjusted, automatically and in real time, to optimize gas-lifting via injection of gas into the well (e.g., using a controller and/or gas-lift optimization system described below).

In one or more embodiments, the gas-lifting for the well site with gas-lift is predicted by a gas-lift optimization system using a machine learning (ML) model, based on (or accepting as inputs), at least, the set of gas-lift parameters. In one or more embodiments, the predicted gas-lifting may include a prediction of the amount of gas-lift obtained (e.g., measured in volume or as a rate) per unit gas injected (or per unit time) as a function of injected depth. In one or more embodiments, the ML model may further make use of gas-lift data obtained from both the well and the reservoir. The gas-lift data may include the reservoir temperature and pressure, the wellhead pressure, the gas-to-oil ratio, and the end of tubing depth, in addition to other information related to the well site with gas-lift such as the configuration of the well. Using the ML model to predict the gas-lifting, an optimal set of gas-lift parameters may be determined by applying an optimizer to the ML model.

FIG. 1 shows a schematic diagram in accordance with one or more embodiments. More specifically, FIG. 1 illustrates a well site with gas-lift (101) including a well (115). A gas-lift system “lifts” production fluids (103), such as oil, gas, and/or water, to a well exit (105) by lowering the density of the production fluids (103) with high pressure gas (107). In one or more embodiments, the gas (107), commonly nitrogen, is pumped from a gas source (131) using a gas pump (133) as part of a gas-lift system and is supplied into the well (115) through a surface gas injection valve (109). In some embodiments, a gas-lift system may simply include one gas pump, though in other embodiments the gas-lift system may include several gas pumps. A gas-lift system further includes electronic controllers (not shown) which may be operated by a computer, such as the computer of FIG. 7 and its accompanying description. The gas source (131) can be of many shapes and sizes. In one or more embodiments, the gas source (131) and gas-lift system including a gas pump (133) are located near the well site (101), as illustrated in FIG. 1. In other examples, the gas source (131) and gas pump (133) may be located at some distance from the well site (101). A surface gas injection valve (109), disposed at a surface (111) and connected to an annulus (113) of the well (115), controls the flow of the gas (107) and injects the gas (107) into the annulus (113) with a user-defined pressure, which can be considered as high (e.g., 20 MPa), according to one or more embodiments. The annulus (113) consists of the space between a casing (117) and a main inner pipe (119) and is configured to isolate the gas (107), allowing the gas (107) to flow the depth of the well (115) without mixing with other fluids. In one or more embodiments, the gas may instead be injected using coiled tubing threaded into the main inner pipe (119) instead of injecting the gas through the annulus (113). Gas injection may involve defining the depth at which the gas is injected, e.g., into the main inner pipe (119), to lift the fluids. For example, injecting gas with coiled tubing may involve defining the depth of the coil tubing landing for gas injection. The depth at with gas is injected in a gas-lift operation can affect the efficiency of the gas-lift operations and the quantity and rate of hydrocarbons extracted from the well.

The casing (117), disposed in the well (115) against a wellbore (121) and typically formed of a durable material such as steel, extends to a depth above a reservoir (123). The casing (117) isolates the subsurface fluids and supports a wellbore (121), a drilled hole comprising the well (115), up to the depth above the reservoir (123). At the other end, the wellbore (121) extends through the reservoir (123) beneath the casing (117). The reservoir (123) is disposed below the surface (111) of the Earth in porous rock formations and is the source of the production fluids (103). The main inner pipe (119), disposed in the wellbore (121), extends from a well exit (105) to a depth proximate or intersecting the reservoir (123) and is a conduit for production fluids (103) to exit the well. In one or more embodiments, the main inner pipe (119) may be formed of tempered steel or equivalent. For the embodiment depicted in FIG. 1, the production fluids (103) may be oil and gas. The oil and gas flows from the reservoir (123) into the wellbore (121) and into the main inner pipe (119). If the reservoir (123) has enough pressure, the oil travels upwards in the main inner pipe (119) to the surface (111). Conversely, if the reservoir (123) does not have enough pressure to lift the oil and gas by itself, in a gas-lift system, external gas (107) is injected into the well (115).

In one or more implementations, at a given pressure that may be considered as high (e.g., 20 MPa), the gas (107) enters the main inner pipe (119) through a downhole gas injection valve (127) and mixes with the production fluids (103) disposed in the main inner pipe (119). The downhole gas injection valve (127) forms a hydraulic connection between the annulus (113) and the main inner pipe (119) and is configured to allow the flow of the gas (107) from the annulus (113) into the main inner pipe (119). In one or more embodiments, the downhole gas injection valve (127) forms a hydraulic connection between coil tubing which is disposed within the main inner pipe (119) and is configured to allow the flow of gas (107) from the coil tubing into the main inner pipe (119). In one or more embodiments, the downhole gas injection valve (127) is composed of stainless steel or the equivalent. The pressure of the injected gas (107), combined with the lighter weight of the gas (107), lowers the density of the production fluids (103) until the mixture becomes light enough to flow towards the well exit (105) and into a production tank (129). The production tank (129) is a storage tank, disposed at the surface (111), that collects and stores the production fluids (103) after the production fluids (103) exit the well (115). In some instances, the production fluids (103) may be directly transported to (e.g., by a pipeline), or otherwise processed by, an oil and gas processing facility.

In accordance with one or more embodiments, one or more downhole sensors (135) are disposed within the well (115). The one or more downhole sensors (135) measure downhole fluid and process properties such as gas flow rate, downhole gas injection rate, downhole liquid production rate, downhole pressure, downhole temperature, and downhole gas concentration. In one or more embodiments, the downhole sensors (135) are grouped together and located on the side of the wellbore (121), as illustrated in FIG. 1. In other examples, the downhole sensors (135) may be separated and located anywhere below the surface (111). Downhole sensors (135) may further be disposed within the reservoir (123) and measure properties of the reservoir such as reservoir temperature and reservoir pressure.

In accordance with one or more embodiments, one or more surface sensors (137) are disposed at, or above, the surface (111) of the well (115). The one or more surface sensors (137) measure fluid and process properties at the surface (111) such as surface pressure, gas injection pressure, and gas injection temperature. In one or more embodiments, the surface sensors (137) are grouped together and located in the vicinity of the well exit (105), as illustrated in FIG. 1. In other examples, the surface sensors (137) may be separated and located anywhere at or above the surface (111). In one or more embodiments, the downhole sensors (135) or surface sensors (137) further include an acoustic sensor, such as a geophone, acquiring acoustic data that is added to the well data. In one or more embodiments, the downhole sensors (135) or surface sensors (137) further include at least one contamination sensor, configured to detect a presence, and possibly a concentration of contaminant in a produced hydrocarbon. The data measured by the downhole sensors (135) and surface sensors (137) are recorded and/or stored using a data collection unit (139). In one or more embodiments, the data collection unit (139) is located in the vicinity of the well exit (105), as illustrated in FIG. 1. In other examples, data collection unit (139) may be located separate from the well but in communication using either wired relays or a remote signal.

Configuring a gas-lift system (including, for example, the operating conditions of a gas pump (133)) for optimal gas-lift is a difficult and laborious task. Further, the state and behavior of a well site with gas-lift (101) may be transient according to conditions at the environment of the site (e.g., changing weather conditions in an offshore well) or within the reservoir (e.g., seismic activity and changing temperature and pressure). Gas-lift stability, in particular, is challenging to maintain. Gas-lift stability refers to the ability of a gas-lift system to maintain consistent and efficient operations over time without experiencing issues that could disrupt production. Instabilities in gas-lift operations can lead to reduced production rates, increased operational costs, and potential damage to equipment.

A notable example of gas-lift instability is a slug flow, in which large pockets of gas and liquid alternate in the production tubing, causing rapid changes in pressure and flow rates. A slug flow may lead to inefficient lifting, decreased production, and wear and tear on well equipment. The gas injection rate is typically balanced in order to create sufficient buoyancy to lift the fluids without causing excessive gas breakout at the surface. Over-injecting gas might lead to inefficient gas usage, while under-injecting gas could result in poor production rates. The gas-to-liquid ratio may also play a significant role in gas-lifted hydrocarbon production, as injecting too much gas may lead to inefficient fluid separation at the surface, while injecting too little gas may result in poor lifting performance. The opening and closing pressures of the surface gas injection valve (109) and the downhole gas injection valve (127) determine when gas injection starts and stops. Properly setting these pressures may help maintain an effective gas-lift cycle. Last, it may be advantageous, in some embodiments, for the hydrocarbon production rate itself to be adjusted according to different circumstances. For instance, the hydrocarbon production rate may be intentionally reduced while a maintenance action is performed. As another example, the hydrocarbon production rate may be desired to change based on current demand.

Though FIG. 1 depicts a well that may be considered typical for onshore wells, the elements shown and described are also generally applicable to offshore (and “rigless”) wells. Consequently, FIG. 1 should not be considered as limiting as embodiments disclosed herein can be readily applied to offshore and rigless well sites although not depicted.

Regarding the components of a well site with gas-lift, the various systems and facilities may each have associated data that describe the status of the given system or facility. Additionally, the systems and/or facilities of a well site with gas-lift can have associated parameters that define and control one or more aspects of their operation. For example, features of the well site with gas-lift may be recorded in measurements collectively referred to as “gas-lift data.” In one or more embodiments, the gas-lift data may include “well data” from the well and “reservoir data” from the reservoir. The well data may describe both static properties of the well (i.e., properties generally invariant to the passage of time), such as its diameter, end of tubing depth, perforation depth, and geographical location (among others not listed) and dynamic properties of the well (i.e., properties that can fluctuate/change temporally) such as the wellhead pressure, gas-to-oil ratio, fluid density inside the well, fluid temperature inside the well, fluid pressure inside the well, water cut, and liquid flowing rate (among others not listed). The reservoir data may describe both static properties of the reservoir, such as its depth, extent, and geographical location (among others not listed) and dynamic properties of the reservoir such as its temperature, pressure, salinity, porosity, resistivity, radioactivity, and fluid density inside the reservoir (among others not listed).

In one or more embodiments, the external gas source and gas pump (and other elements, components, and or facilities of a gas-lift system) have an associated set of gas-lift parameters. The set of gas-lift parameters may include gas injection parameters that define the aspects of the gas injection into the well. The gas injection parameters may include defining a rate of gas injection, a fluid conduit medium (e.g., coil tubing or tubing annulus), and a depth at which gas is to be injected. To clarify, the selection or identification of the element(s) through which gas is injected into the well are referenced herein as a “fluid conduit medium,” and examples of fluid conduit mediums include coiled tubing and using an annular volume in the wellbore around the main pipe, i.e., a tubing annulus. A volume of injected gas may be defined by the injection parameters as well, though it is noted that a volume may be considered equivalent with a rate through integration of the rate over time.

In one aspect, embodiments disclosed herein relate to a system for determining the set of gas-lift parameters that optimize the gas-lifting at a well site with gas-lift. Optimizing the gas-lifting may refer to several goals, depending on the circumstances of the well site with gas-lift. For example, in one or more embodiment, optimizing gas-lifting may include determining the set of gas-lift parameters such that the liquid flowing rate of fluid production is maximized, while in other embodiments, optimizing gas-lifting may include determining the set of gas-lift parameters such that the least amount of injected gas is utilized to achieve production from the well. In some embodiments, optimizing the gas-lifting may also refer to determining the set of gas-lift parameters such that both the liquid flowing rate of fluid production is maximized while the amount of injected gas is minimized, or achieving the maximum gas-lift per unit gas injected. In one or more embodiments, optimizing the gas-lifting may also refer to determining the maximum gas-lift per unit gas injected subject to a specific constraint, such as forcing the injection depth to be the end of tubing depth (which may be applicable, for example, in a well-cleaning operation).

In one or more embodiments, an optimal set of gas-lift parameters are determined with a machine learning (ML) model, taking into consideration the current state of the well via the well data, the current state of the reservoir via the reservoir data, and possible configurations of the external source of gas and gas pump via the set of gas-lift parameters (or, more generally, the possible configurations of an associated gas-lift system). In accordance with one or more embodiments, the set of gas-lift parameters may be adjusted automatically, and in real-time, through control systems disposed throughout the various elements (e.g., the gas pump may be controlled by a gas pump control system) capable of receiving commands, the commands based on the ML model under optimization.

In accordance with one or more embodiments, gas-lift data (including, in some embodiments, well data from a well and reservoir data from a reservoir) are processed with a ML model to predict the gas-lifting that may be achieved as a result of injecting gas into the well located at a well site with gas-lift. The predicted gas-lifting may be constructed in a variety of forms. In one or more embodiments, the predicted gas-lifting may include a prediction of the amount of gas-lift obtained (e.g., measured in volume or as a rate) per unit gas injected (or per unit time) as a function of injected depth. When coil tubing is used as the flowing medium, the injected depth may be referred to as the coil tubing landing depth. The predicted gas-lifting is used to determine an optimal set of gas-lift parameters, in addition to a prediction of the liquid flowing rate of production fluid considered at a well site with gas-lift upon updating the set of gas-lift parameters to the optimal set of gas-lift parameters.

Machine learning (ML), 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 machine learning (ML), 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.

Machine learning (ML) model types may include, but are not limited to, neural networks, decision trees, random forests, support vector machines, generalized linear models, and Bayesian regression. ML 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 ML model, is to determine the gas-lifting from a well site with gas-lift in view of gas-lift data from associated structures such as a well and a reservoir (e.g., the gas-to-oil ratio inside a well with access to a reservoir at a certain temperature) that may affect gas-lifting operations when injecting gas into the well. In some embodiments, the ML model is further informed by the set of gas-lift parameters defining the operation of an external source of gas such as a gas-lift system including a gas pump.

FIG. 2 depicts one embodiment of a well site with gas-lift (e.g., well site with gas-lift B (299)). Owing to their similarity, the embodiments of FIG. 2 and FIG. 1 may both be referred to as “well sites with gas-lift,” though they need not be identical. As such, similar elements are present but are given distinct labels for convenience. The well site with gas-lift (e.g., well site with gas-lift B) includes a well (200) with access to a reservoir (206). In this particular embodiment, a gas-lift system (203) both houses and is capable of injecting gas into the well (200). In one or more embodiments, the gas-lift system (203) includes a gas pump. A commonly used gas for gas-lifting is nitrogen, though in other embodiments, alternative sources of gas and alternative mechanisms for injecting gas other than a gas pump may be considered. The well (200) may be either onshore or offshore, and in association with a rig or without a rig.

For clarity, with regard to the instant disclosure, properties of elements of a well site with gas-lift that are not generally considered configurable are referred to as “data” while features and aspects of elements of a well site with gas-lift that are generally considered configurable are referred to as “parameters.” As seen in FIG. 2, the well (200) has associated with itself well data (218) and the reservoir (206) has associated with itself reservoir data (230). The gas-lift system (203) has associated with itself gas injection parameters, which define the operation of the gas-lift system (203) and define the properties of the injected gas (221) into the well (200).

The state of the well (200) is described, at least in part, by well data (218). In one or more embodiments, the well data (218) may include information related to both static and dynamic properties of the well. For example, static properties of the well (200) included by the well data (218) may be the well diameter, end of tubing depth, perforation depth, and geographical location. Dynamic properties of the well (200) included by the well data (218) may be the wellhead pressure, gas-to-oil ratio, fluid density inside the well, fluid temperature inside the well, fluid pressure inside the well, water cut, and liquid flowing rate. In one or more embodiments, the reservoir (206) is described, at least in part, by reservoir data (230). The reservoir data (230) may include information related to both static and dynamic properties of the reservoir (206). For example, static properties of the reservoir (206) included by the reservoir data (230) may be its depth, extent, and geographical location. Dynamic properties of the reservoir (206) included by the reservoir data (230) may be its temperature, pressure, salinity, porosity, resistivity, radioactivity, and fluid density. A person of ordinary skill in the art will recognize that other properties of the well (200) and information related to its state not listed here may be included as well data (218), and that other properties of the reservoir (206) and information related to its state not listed here may be included as reservoir data (230). In addition, the properties need not be distinctly labeled as either static or dynamic, and these labels are provided solely to further clarify what may be considered as well data (218) and reservoir data (230).

In one or more embodiments, the operation of the gas-lift system (203) and the properties of the injected gas (221), from the gas-lift system (203) into the well (200), are defined by gas injection parameters (227). The gas injection parameters (227) may include defining aspects of the injection operation such as the fluid flowing medium (e.g., coil tubing, or tubing annulus), and defining aspects of the injected gas (221) such as the injection rate (or volume) and the pressure and temperature of the injected gas (221). The gas injection parameters (227) may further include defining the injection depth. In summary, the gas injection parameters (227) include defining various aspects related to injecting gas into the well (200) for gas-lifting procedures, and a person of ordinary skill in the art will understand that the parameters listed here are not exhaustive and others may be included in one or more embodiments of the instant disclosure.

In accordance with one or more embodiments, FIG. 2 shows that the data relevant to gas-lift well procedures are collectively stored as gas-lift data (209). From the well (200) this may include the well data (218) and the reservoir data (230). The well data (218) may include information regarding the well head pressure and the oil-to-gas ratio in the well (200). The reservoir data (230) may include information regarding the temperature or pressure inside the reservoir (206). Similarly, the parameters that define the operation of configurable aspects of the well site with gas-lift are collectively stored as a set of gas-lift parameters (212). In the embodiment of a well site with gas-lift depicted by FIG. 2 (e.g., well site with gas-lift B (299)), this may include the gas injection parameters (227). The gas injection parameters may include defining the volume of gas injected into the well. A commonly used gas for gas-lifting is nitrogen, though in other embodiments, alternative sources of gas may be considered. A person with ordinary skill in the art will appreciate that additional gas-lift data (i.e., data relevant to gas-lift procedures) may be collected beyond the listed examples, and that additional gas-lift parameters (i.e., configurable aspects of the well site with gas-lift) may be considered without departing from the scope of this disclosure.

In accordance with one or more embodiments, the gas-lift data (209) and the set of gas-lift parameters (212) are processed by an ML model (233) as part of a gas-lift optimization system (215). In one or more embodiments, the result of the processing by the ML model (233) is a prediction of the gas-lifting (238). The predicted gas-lifting may be constructed in a variety of forms. In one or more embodiments, the predicted gas-lifting may include a prediction of the amount of gas-lift obtained (e.g., measured in volume or as a rate) per unit gas injected (or per unit time) as a function of injected depth. The ratio between volume of fluid produced relative to the volume of gas injected may be referred to as the “gas-lift utilization” (GLU). In one or more embodiments, the GLU is predicted across the length of the wellbore for various injection depths, forming a predicted GLU-depth relation (239). While FIG. 2 depicts the ML model (233) as being applied to a well site with gas-lift (e.g., well site with gas-lift B (299)), herein, the minimum number of wells (200), gas-lift systems (203), and reservoirs (206), required to form a function well site with gas-lift are each one. As such, in one or more embodiments, the ML model (233) is applied to a single well (200), a single gas-lift system (203) and a single reservoir (206). In one or more embodiments, the ML model (233) is applied to a single well (200), a single gas-lift system (203) and a single reservoir (206) even when more than one of each of these systems is included. That is, the ML model (233) may be applied to many combinations of wells (200), gas-lift systems (203), and reservoirs (206) or across individual groupings including only one of each of these systems without limitation.

In accordance with one or more embodiments, the gas-lift data (209), and the set of gas-lift parameters (212) may be pre-processed before being processed by the ML model (233). 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 includes identifying and selecting a subset of gas-lift data (209) with the greatest discriminative power with respect to predicting gas-lift (238). For example, in one embodiment, discriminative power may be quantified by calculating the strength of correlation between elements of the gas-lift data (209) and the predicted quantities (i.e., predicted gas-lifting (238)). Consequently, in some embodiments, not all of the gas-lift data (209) need be passed to the ML model. Feature engineering encompasses combining, or processing, various gas-lift data (209) to create derived quantities. The derived quantities can be processed by the ML model (233). For example, the gas-lift data (209) may be processed by one or more “basis” functions such as a polynomial basis function or a radial basis function. In some embodiments, the gas-lift data (209) is passed to the ML model (233) 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 gas-lifting (238) is used to determine the set of gas injection parameters (212) that optimizes gas-lifting. In one or more embodiments, the set of gas-lift parameters (212) define an operational state of the mechanisms responsible for gas-lifting (e.g., the gas-lift system (203)) and thus affect a well site with gas-lift (e.g., well site with gas-lift B (299)) in its ability to obtain optimal gas-lifting (which may vary depending on the goal of the gas-lifting procedure). Determining the optimal set of gas-lift parameters is carried out by an optimizer (236) applied to the ML model (233), as part of the gas-lift optimization system (215). The optimizer (236) is described in further detail below and in reference to FIG. 5. In short, the optimizer (236) may include an algorithm or collection of algorithms, without limitation, that use the predictive capabilities of the ML model (233) to explore the parameter space spanned by the ML model and its inputs (e.g., the gas-lift data (209) and the set of gas lift parameters (212)) in relation to its outputs (i.e., the predicted gas-lifting (238)).

In one or more embodiments, the GLU is predicted across the length of the wellbore for various injection depths, forming a predicted GLU-depth relation (239). The predicted GLU-depth relation (239), put simply, predicts the volume of produced fluid relative to the volume of gas injected at different injection depths. As such, the predicted GLU-depth relation (239) may be used, by the optimizer (236), to determine the optimal injection depth (242) and the optimal injection rate (243) for gas injection during gas-lifting.

As previously established, injection depth and injection rate may each be parameters considered among the set of gas injection parameters (227) and therefore the set of gas-lift parameters (212). Accordingly, the optimizer (236) as applied to the ML model (233), as part of the gas-lift optimization system (215) may further output a predicted optimal liquid rate (245) of fluid production, which is the liquid rate of fluid production that will be achieved by applying the optimal parameters (i.e., the optimal injection depth (242) and the optimal injection rate (243)) across the well site with gas-lift (e.g., well site with gas-lift B (299)).

As previously established, in one or more embodiments, the set of gas-lift parameters includes gas injection parameters. The gas injection parameters may include defining the depth of gas injection for gas-lifting, and the rate of gas injection for gas-lifting, among other properties of the gas-injecting procedure. FIG. 3 depicts one possible intermediate output of the ML model (233) under optimization (236). It is emphasized that this depiction is used only to illustrate how the optimal set of gas-lift parameters may be determined and other embodiments may differ from the depiction included in FIG. 3. More specifically, FIG. 3 depicts (for a given well (200), gas-lift system (203), and reservoir (206)) the predicted relationship between the liquid rate of production fluid and injection depth for various gas injection rates. Further depicted in FIG. 3 is the predicted GLU-depth relation (239) for gas-lift utilization (GLU; defined above as the volume of produced fluid relative to the volume of injected gas as a function of injection depth) at various gas injection rates. The vertical axis indicates the injection depth with deeper values, with respect to the Earth's surface (or values representing distances further into the wellbore) closer to the origin (or the intersection, at the bottom left of the figure) of the axes. The horizontal axis indicates either the liquid rate (indicated by the top horizontal axis) or the GLU factor (indicated by the bottom horizontal axis) with increasing liquid rate and increasing GLU further from the origin (or intersection, at the bottom left of the figure) of the axes. In one or more embodiments, FIG. 3 can be described as a “lift capability plot.”

For illustration, consider the six curves shown in the lift capability plot of FIG. 3. A first liquid rate curve (300) is depicted indicating the liquid rate achieved (top horizontal axis) when injecting gas at a rate of 300 scf/min at particular depths (vertical axis). Additionally, a first GLU curve (305) is depicted indicating the GLU factor achieved (bottom horizontal axis) when injecting gas at a rate of 300 scf/min at particular depths (vertical axis). A second liquid rate curve (310) is depicted indicating the liquid rate achieved when injecting gas at a rate of 400 scf/min at particular depths. Additionally, a second GLU curve (315) is depicted indicating the GLU factor achieved when injecting gas at a rate of 400 scf/min at particular depths. A third liquid rate curve (320) is depicted indicating the liquid rate achieved when injecting gas at a rate of 500 scf/min at particular depths. Additionally, a third GLU curve (325) is depicted indicating the GLU factor achieved when injecting gas at a rate of 500 scf/min at particular depths. As can be seen by first liquid rate curve (300), second liquid rate curve (310), and third liquid rate curve (320), the liquid rate of produced fluid the system generally tends to increase for increasing gas injection rates. However, the gas injection rate that yields the greatest liquid rate may not be the most efficient per unit gas injected, i.e., the gas injection rate may not achieve the maximum GLU. For example, the injection rate that achieves the maximum liquid rate is given by the third liquid rate curve (320), as the third liquid rate curve (320) achieves a value along the horizontal axis greater than the maximum values of both the first liquid rate curve (300) and the second liquid rate curve (310). However, the injection rate that achieves the maximum GLU is given by the first GLU curve (305), as the first GLU curve (305) achieves a value along the horizontal axis greater than the maximum values of both the second GLU curve (315) and the third GLU curve (325). According to FIG. 3, an injection rate of 300 scf/min is the most efficient (or, achieves the greatest GLU), as shown by the first GLU curve (305), despite the fact that an injection rate of 500 scf/min achieves the greatest overall liquid rate of produced fluid, as shown by the third liquid rate curve (320).

In one or more embodiments, as depicted in FIG. 3, the set of gas-lift parameters, including the gas injection parameters, includes the depth of gas injection for gas-lifting and the rate of gas injection for gas-lifting. Depending on the goal at deployment of the ML model (233), it may be advantageous to maximize either the liquid rate or the GLU, among other possible quantities. For illustration, the latter case of maximizing GLU is considered. In embodiments that include the gas injection depth and the gas injection rate as individual parameters in the set of gas-lift parameters, the optimum values may be indicated by the GLU curve that reaches furthest along the horizontal axis in a lift capability plot, such as FIG. 3. As discussed above, the curve achieving the maximum GLU is the first GLU curve (305), which corresponds to an injection rate of 300 scf/min. Accordingly, In FIG. 3, the optimal injection depth is indicated by the label “optimum, stage 1” (330). In one or more embodiments, a second set of optimal parameter values may be determined. In one scenario, if a well requires cleaning, it may be advantageous to inject gas at the bottom of the well (e.g., at the end of tubing depth). Alternatively, if the static fluid levels in the well are not well measured, it may be advantageous to again inject gas at the bottom of the well to increase the probability of success of the gas-lifting operation. In either case, the optimum injection rate may be determined with the injection depth set to a value corresponding to the end of tubing depth (or bottom of the well). The optimal parameter values corresponding to injection at the bottom of the well are indicated by the label “optimum, stage 2” (340). In this case, either the first GLU curve (305) corresponding to an injection rate of 300 scf/min, or the second GLU curve (315) corresponding to an injection rate of 400 scf/min, may be considered optimal, as their maximum GLU are similar to each other, and both first GLU curve (305) and second GLU curve (315) achieve greater GLU than the third GLU curve (325) corresponding to an injection rate of 500 scf/min.

Returning to FIG. 2, in one or more embodiments, the ML model (233), as part of the gas-lift optimization system (215) outputs the predicted gas-lifting (238) in a format compliant with other software for visualization or further processing. In one or more embodiments, the output (238) of the ML model (233) and optimizer (236) is digitally formatted in such a way to be readily used, without further modification, with a Microsoft Excel® Macro Visual Basic Application (VBA). In one or more embodiments, the output of the ML model (233) and optimizer (236) is digitally formatted in such a way to be readily used, without further modification, with the well and pipeline modeling code PROSPER. As previously established, injection depth and injection rate may each be parameters considered among the set of gas injection parameters (227) and therefore the set of gas-lift parameters (212). Accordingly, the optimizer (236) as applied to the ML model (233), as part of the gas-lift optimization system (215) may further output a predicted optimal liquid rate (245) of fluid production, which is the liquid rate of fluid production that will be achieved by applying the optimal parameters (i.e., the optimal injection depth (242) and the optimal injection rate (243)) across the well site with gas-lift (e.g., well site with gas-lift B (299))

Upon identifying the set of gas-lift parameters that optimize gas-lifting (i.e., the optimal set of gas-lift parameters), this set of parameters may be applied across the well site with gas-lift (e.g., well site with gas-lift B (299)) and associated components automatically using control systems that receive commands from a control system, such as a gas-lift optimization system (215). The gas-lift optimization system (215) may be located anywhere within the well site with gas-lift (e.g., well site with gas-lift B (299)) including near to the well (200) or near to the gas-lift system (203). In other implementations, the gas-lift optimization system (215) is located remotely relative to the well site with gas-lift (e.g., using a cloud computing system). The command issued from the gas-lift optimization system (215), being a command to apply the new set of optimal parameters, is indicated by Command X (248) in FIG. 2. In one or more embodiments, gas-lifting is continuously monitored and recorded by at least one downhole and surface sensor disposed within the well to ensure that the accepted optimal set of gas-lift parameters maintain the well site with gas-lift (e.g., well site with gas-lift B (299)) at optimal gas-lifting. That is, the accepted set of gas-lift parameters are validated by measuring the gas-lifting and determining whether the gas-lifting is improved after adjusting the set of gas-lift parameters (212) to the optimal set of gas-lift parameters.

The ML model (233) may be of any type known in the art. Generally, the ML model type and architecture with the greatest performance on a set of hold-out data is selected. Training an ML model involves processing data to develop a functional relationship between the inputs and the targets of the modelling data. The trained ML model may be described as a function relating the inputs and the outputs. That is, the ML model may be mathematically represented as outputs=ƒ(inputs), such that given an input the ML model (233) may produce an output. To better explain the workings of the ML model (233) and concepts related to machine learning and training more generally, consider one possible embodiment where the ML model (233) is an artificial neural network. A diagram of a neural network is shown in FIG. 4. At a high level, a neural network (400) may be graphically depicted as being composed of nodes (402), where here any circle represents a node, and edges (404), shown here as directed lines. The nodes (402) may be grouped to form layers (405). FIG. 4 displays four layers (408, 410, 412, 414) of nodes (402) where the nodes (402) are grouped into columns, however, the grouping need not be as shown in FIG. 4. The edges (404) connect the nodes (402). Edges (404) may connect, or not connect, to any node(s) (402) regardless of which layer (405) the node(s) (402) is in. That is, the nodes (402) may be sparsely and residually connected. A neural network (400) will have at least two layers (405), where the first layer (408) is considered the “input layer” and the last layer (414) is the “output layer.” Any intermediate layer (410, 412) is usually described as a “hidden layer.” A neural network (400) may have zero or more hidden layers (410, 412) and a neural network (400) with at least one hidden layer (410, 412) may be described as a “deep” neural network or as a “deep learning method.” In general, a neural network (400) may have more than one node (402) in the output layer (414). In this case the neural network (400) may be referred to as a “multi-target” or “multi-output” network.

Nodes (402) and edges (404) carry additional associations. Namely, every edge is associated with a numerical value. The edge numerical values, or even the edges (404) themselves, are often referred to as “weights” or “parameters.” While training a neural network (400), numerical values are assigned to each edge (404). Additionally, every node (402) is associated with a numerical variable and an activation function. Activation functions are not limited to any functional class, but traditionally follow the form:

$\begin{array}{cc}A=f\left({\sum }_{i\in \left(\mathrm{incoming}\right)}\left[{\left(\mathrm{node}\mathrm{value}\right)}_i{\left(\mathrm{edge}\mathrm{value}\right)}_i\right]\right),& \left(1\right)\end{array}$

• • where i is an index that spans the set of “incoming” nodes (402) and edges (404) and f is a user-defined function. Incoming nodes (402) are those that, when viewed as a graph (as in FIG. 4), have directed arrows that point to the node (402) where the numerical value is being computed. Some functions for ƒ may include the linear function ƒ(x)=x, sigmoid function

$f\left(x\right)=\frac{1}{1+e^{-x}},$

and rectified linear unit function ƒ(x)=max (0, x), however, many additional functions are commonly employed. Every node (402) in a neural network (400) 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 (400) receives an input, the input is propagated through the network according to the activation functions and incoming node (402) values and edge (404) values to compute a value for each node (402). That is, the numerical value for each node (402) may change for each received input. Occasionally, nodes (402) are assigned fixed numerical values, such as the value of 1, that are not affected by the input or altered according to edge (404) values and activation functions. Fixed nodes (402) are often referred to as “biases” or “bias nodes” (406), displayed in FIG. 4 with a dashed circle.

In some implementations, the neural network (400) may contain specialized layers (405), 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 (400) comprises assigning values to the edges (404). To begin training the edges (404) 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 (404) values have been initialized, the neural network (400) 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 (400) 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.

Using the neural network as an example, and in the context of the instant disclosure, the input of the neural network is the gas-lift data (which may be pre-processed), and the set of gas-lift parameters, while the target is gas-lifting. Gas-lifting, as a target output (or prediction), may be constructed in a variety of forms. In one or more embodiments, the output gas-lifting may include the amount of gas-lift obtained (e.g., measured in volume or as a rate) as a result of the injection operation. Further, the output gas-lifting may be considered relative to the volume of gas injected (or per unit time) as a function of injected depth. The neural network (400) output is compared to the associated input data target(s). However, for a given well, these outputs may not be known ahead of time, and the neural network must be trained using alternative training data (inputs and targets). The same input-output structure can be used for other supervised machine-learned model types. That is, while a neural network is provided as an example, it is emphasized that the ML model (233) need not be a neural network. Other machine-learned models types such as a random forest or a gradient boosting machine can be readily trained using the target and inputs described herein, in accordance with one or more embodiments.

In one or more embodiments, training inputs and targets may come from empirical well data including a plurality of histories of well behavior as measured and recorded from analogous and substantially similar wells. The plurality of histories of well behavior may include substantially similar gas-lift data and gas-lift parameters as the embodiments of well sites with gas lift of the instant disclosure. The training gas-lift data may include both well data and reservoir data. The well data may include information regarding the well head pressure, the oil-to-gas ratio in the well, the end of tubing depth, the water cut level, and the completion fluid density and viscosity, and the top of perforation depth, among other measurements not listed. The reservoir data may include information regarding the temperature and pressure inside the reservoir, as well as the reservoir depth, among other measurements not listed. The training gas-lift parameters may include historical gas injection parameters, the gas injection parameters defining aspects of the injection operation such as the fluid flowing medium (e.g., coil tubing, or tubing annulus), and defining aspects of the injected gas such as the injection rate (or volume), the pressure and temperature of the injected gas, and the injection depth. The plurality of histories of well behavior may further include the necessary training targets, that is, the achieved gas-lifting according to the set of gas-lift parameters and in view of the gas-lift data.

In one or more embodiments, the training data may further include simulated training data (inputs and targets). Regarding a well site with gas-lift, the simulations may describe the behavior of an oil and gas well according to substantially similar gas-lift data and substantially similar gas-lift parameters, where “similar” is in reference to real well site with gas-lifts. Just as with the empirical well data including a plurality of histories of (real) well behavior, the simulated gas-lift well data used for training may include both simulated well data and simulated reservoir data. The simulated well data may include information regarding the well head pressure, the oil-to-gas ratio in the well, the end of tubing depth, the water cut level, and the completion fluid density and viscosity, and the top of perforation depth, among other measurements not listed. The simulated reservoir data may include information regarding the temperature and pressure inside the reservoir, as well as the reservoir depth, among other measurements not listed. The simulated gas-lift parameters used for training may include simulated gas injection parameters, the gas injection parameters defining aspects of the injection operation such as the fluid flowing medium (e.g., coil tubing, or tubing annulus), and defining aspects of the injected gas such as the injection rate (or volume), the pressure and temperature of the injected gas, and the injection depth. In one or more embodiments, the simulation is created using the software PROSPER, which is capable of simulating oil and gas wells according to the specified positions above. The simulations may further include the necessary training targets, that is, the achieved gas-lifting according to the set of gas-lift parameters and in view of the gas-lift data.

The comparison of the neural network (400) 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 (400) output and the associated target(s). The loss function may also be constructed to impose additional constraints on the values assumed by the edges (404), 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 (404) values to promote similarity between the neural network (400) output and associated target(s) over the data set. Thus, the loss function is used to guide changes made to the edge (404) 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 (404) values. The gradient indicates the direction of change in the edge (404) values that results in the greatest change to the loss function. Because the gradient is local to the current edge (404) values, the edge (404) 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 (404) values or previously computed gradients. Such methods for determining the step direction are usually referred to as “momentum” based methods.

Once the edge (404) values have been updated, or altered from their initial values, through a backpropagation step, the neural network (400) will likely produce different outputs. Thus, the procedure of propagating at least one input through the neural network (400), comparing the neural network (400) output with the associated target(s) with a loss function, computing the gradient of the loss function with respect to the edge (404) values, and updating the edge (404) 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 (404) 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 (404) values are no longer intended to be altered, the neural network (400) is said to be “trained.”

In one or more embodiments, the ML model is a random forest. Generally, a random forest may be considered an ensemble of decision trees. A decision tree is composed of nodes. A decision is made at each node such that data present at the node are segmented. Typically, at each node, the data at said node, are split into two parts, or segmented bimodally, however, multimodal segmentation is possible. The segmented data can be considered another node and may be further segmented. As such, a decision tree represents a sequence of segmentation rules. The segmentation rule (or decision) at each node is determined by an evaluation process. The evaluation process usually involves calculating which segmentation scheme results in the greatest homogeneity or reduction in variance in the segmented data. However, a detailed description of this evaluation process, or other potential segmentation scheme selection methods, is omitted for brevity and does not limit the scope of the present disclosure.

Further, if at a node in a decision tree, the data are no longer to be segmented, that node is said to be a “leaf node.” Commonly, values of data found within a leaf node are aggregated, or further modeled, such as by a linear model. A decision tree can be configured in a variety of ways, such as, but not limited to, choosing the segmentation scheme evaluation process, limiting the number of segmentations, and limiting the number of leaf nodes. Generally, when the number of segmentations or leaf nodes in a decision tree is limited, the decision tree is said to be a “weak learner.” In most implementations, the decision trees from which a random forest is composed are weak learners. Additionally, for a random forest, the decision trees operate independently, and the results are gathered in the end. For each tree, a selection from the training data is randomly made with replacement, and the tree is built upon this selection in a training process known in the art. Generally, the operation of selection with replacement is known as bootstrap aggregating or “bagging.” Further, in some implementations, at each proposed segmentation a random subset of the features is selected to avoid biasing segmentation on only a few features which outperform other features in prediction. However, in some implementations, all features are considered at each proposed segmentation or a subset is selected through a non-random process. For regression, the result of a random forest may be the average prediction from many individual decision trees, whereas for classification, the result of a random forest may be the classification determined by the most trees.

One with ordinary skill in the art will appreciate that many adaptions may be made to the random forest and that these adaptions do not exceed the scope of this disclosure. Some adaptions may be algorithmic optimizations, efficient handling of sparse data, use of out-of-core computing, and parallelization for distributed computing. In additional, though multiple embodiments using different ML models have been suggested, one skilled in the art will appreciate that this process, of determining the set of gas-lift parameters that optimize gas-lifting, is not limited to the listed ML models. ML models such as gradient boosted trees, recurrent neural networks, 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 optimizer (e.g., optimizer (236)) uses an optimization algorithm or method is used to invert, or intelligently probe one or more inputs to determine the set of gas-lift parameters that optimizes gas-lifting given the predicted gas-lifting of the ML model (e.g., ML model (233)). The optimizer may include one algorithm or a collection of algorithms, without limitation, that use the predictive capabilities of the ML model to explore the parameter space spanned by the ML model and its inputs (e.g., gas-lift data (209) and set of gas lift parameters (212)) in relation to its outputs (e.g., predicted gas-lifting (238)). A commonly used non-linear optimizer is the genetic algorithm (GA). An overview of the typical steps used in the genetic algorithm (GA) is provided in FIG. 5. The genetic algorithm (500) begins by generating an initial population or multiple populations (502). A population consists of one or more “individuals.” In the context of the genetic algorithm (500), an individual is a single representation, or encoding, of the function parameters over which optimization is to occur. In the case of well site with gas-lift, the parameters over which optimization is to occur include states or values encompassed by set of gas-lift parameters. In this case, an individual is the assignment of values to each parameter while respecting any imposed constraints (e.g., not permitting the gas to be injected at a depth that exceeds the end of tubing depth of the well). The number of individuals in a population, and the method of initially generating individuals, are hyperparameters chosen by the user. When multiple populations are used, commonly referred to as an “island” scheme, the populations need not be initialized using the same set of hyperparameters.

Once a population(s) has been generated (502), the “fitness” of every individual in the population(s) is evaluated (504). For example, in the context of the well site with gas-lift described in FIG. 2, the parameter values (in this case, the set of gas-lift parameters (212)) of an individual are used to predict a gas-lifting scenario. The predicted gas-lifting may be constructed in a variety of forms. In one or more embodiments, the predicted gas-lifting may include the volume of fluid produced relative to the volume of gas injected. In one or more embodiments, the objective may be to maximize the volume of fluid produced relative to the volume of gas injected, and under these circumstances, individuals that result in greater volumes of produced fluid per unit gas injected are considered “more fit.” In other embodiments, the objective may simply be to maximize the amount of fluid produced, and under these circumstances, individuals that result in greater volumes of produced fluid are considered “more fit.”

Next, a stopping criterion is checked (506). Many stopping criteria exist, including, but not limited to, the number of iterations the genetic algorithm has run, the maximum or minimum fitness score achieved by an individual, the relative change in fitness scores between iterations, and the similarity of individuals in a population, or combinations of these criteria. If the algorithm is to stop, typically, the most fit individual(s) seen during the genetic algorithm process is selected (512) and the algorithm terminates. Likewise, if the genetic algorithm continues, one or more individuals from the population(s) are selected (508). This selection may be done by simply selecting the portion of the population with the highest fitness scores, or through a tournament process, or other selection mechanism.

Once individuals have been selected (508), the individuals may be propagated through without alteration, removed, or altered through so-called crossover, mutation, and differential evolution methods to create “offspring” (510). The offspring are themselves individuals; that is, new representations, or encodings, of the function parameters. It is noted that many evolutionary methods exist to create offspring and the preceding list is not all-inclusive and should be considered non-limiting. The offspring are then evaluated for fitness (504) and the process is repeated until the genetic algorithm stopping criterion is met.

Again, the description of the genetic algorithm (GA) provided in FIG. 5 is generalized and one skilled in the art will appreciate that many modifications can be made, and are regularly made, to the genetic algorithm (GA) without departing from its intended scope. For example, an enhanced genetic algorithm (eGA) is developed by including additional features such as: enforcing deliberately diverse initial populations; using heterogenous hyperparameters for each population; dynamically updating or scheduling changes to the selection process and offspring creation process; using a unidirectional migration policy between populations; adding additional checks, such as a premature convergence check; using a self-adaptive differential evolution method; performing a localized exhaustive search in regions of stagnation or saturation.

Other non-linear optimizers may be employed to optimize the predicted gas-lifting and determine the optimal set of gas-lift parameters. The non-linear optimizer could be a Bayesian-based optimizer which elects new parameters based on an analysis of the updated posterior distribution. In this context, the level of exploration and exploitation would be determined by the user. The optimizer, including an optimization algorithm such as the one described above and in FIG. 5, may be used to create a lift capability plot, such as the lift capability plot depicted in FIG. 3. In one aspect, the lift capability plot may be considered the output of a “brute-force” method of optimization commonly referred to as a “grid-search” method. In a brute-force grid-search optimization method, the parameter space (e.g., the parameter space of injection depth and injection rate) is discretized (i.e., into a grid of N dimensions, where N is the number of independent parameters in the parameter space) and a prediction (e.g., the predicted liquid rate of fluid production or the predicted GLU) is made for each possible combination of parameter values (e.g., for each possible combination of injection depth and injection rate). Grid-search techniques are straightforward and ensure that the entirety of the parameter space under consideration is explored. However, grid-search techniques are typically computationally intensive and may spend excessive computational resources considering parameter spaces that are far from optimal, thus motivating alternative optimization routines like the genetic algorithm, Bayesian-based optimizations, and Markov Chain Monte Carlo (MCMC) optimization. The listed optimization routines that may be included by the optimizer in the gas-lift optimization system are provided for illustration and should not be considered limiting with respect to the instant disclosure. Alternative optimization algorithms may be considered, without limitation.

In some embodiments, the optimal sets of parameters, as determined by the optimizer applied to the ML model, are validated by measuring the liquid rate of produced fluid (or the gas-lifting achieved) and determining whether the liquid rate of produced fluid (or the gas-lifting achieved) is improved after adjusting the set of gas-lift parameters to the optimal set of gas-lift parameters.

The process of evaluating gas-lift data and determining the set of gas-lift parameters that optimize gas-lifting at a well site with gas-lift is summarized in the flow chart of FIG. 6. In Block 601, gas-lift data is obtained from a well site with gas-lift, the well site with gas-lift including an oil and gas well with access to a hydrocarbon reservoir and a gas-lift system including a gas pump. The gas-lift data may include well data describing the state of the well as measured from a plurality of surface and subsurface devices disposed throughout the well. The well data may describe both static properties of the well, such as its diameter, end of tubing depth, perforation depth, and geographical location (among others not listed) and dynamic properties of the well such as the wellhead pressure, gas-to-oil ratio, fluid density inside the well, fluid temperature inside the well, fluid pressure inside the well, water cut, and liquid flowing rate (among others not listed). The gas-lift data may also include reservoir data describing the state of the reservoir as measured from at least one subsurface device disposed within the reservoir. The reservoir data may describe both static properties of the reservoir, such as its depth, extent, and geographical location (among others not listed) and dynamic properties of the reservoir such as its temperature, pressure, salinity, porosity, resistivity, radioactivity, and fluid density inside the reservoir (among others not listed).

In one or more embodiments, the gas-lift data are 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 603, gas-lift parameters related to the well site with gas-lift are obtained. The gas-lift parameters may include gas injection parameters defining the operation of injecting gas into the well from a gas-lift system. The gas injection parameters may include defining the rate of gas injection, the fluid conduit medium (e.g., coil tubing or tubing annulus), and the depth at which gas is to be injected. A volume of injected gas may be defined by the injection parameters as well, though it is noted that a volume may be considered equivalent with a rate through integration of the rate over time.

In Block 605, the gas-lift data and the set of gas-lift parameters are processed by a machine learning (ML) model, to predict gas-lifting at a well site with gas-lift. FIG. 2 depicts one embodiment of the input-output structure of the ML model. The ML model outputs a prediction of the gas-lifting, according to the set of gas-lift parameters and in view of the gas-lift data. The predicted gas-lifting may be constructed in a number of ways. In one or more embodiments, the predicted gas-lifting may include a prediction of the amount of gas-lift obtained (e.g., measured in volume or as a rate) per unit gas injected (or per unit time) as a function of injected depth. The ratio between volume of fluid produced relative to the volume of gas injected may be referred to as the “gas-lift utilization” (GLU). In one or more embodiments, the GLU is predicted across the length of the wellbore for various injection depths, forming a predicted GLU-depth relation.

In Block 607, the optimal set of gas-lift parameters are determined by an optimizer applied to the ML model such that predicted gas-lifting is optimized. The optimal gas-lifting may be defined differently according to the circumstances. For example, in one or more embodiments, the optimal gas-lifting may be defined as a gas-lifting that lifts the largest volume of gas without further consideration. In one or more embodiments, the optimal gas-lifting may be defined as a gas-lifting that utilizes the least amount of injected gas to achieve fluid production. In one or more embodiments, the optimal gas-lifting may be defined as a gas-lifting where the maximum volume of produced fluid is lifted per unit volume of gas injected. One possible embodiment of the optimizer, the genetic algorithm (GA), has been described above and depicted in FIG. 6. Other optimizers may be employed, such as a Bayesian-based optimizer, without limitation.

In Block 609, the set of gas-lift parameters are adjusted (e.g., through interactions between control systems located within the well site with gas-lift) to their optimal values as determined using the ML model and optimizer. This adjustment may be performed automatically and autonomously, or may be done manually, or may be checked by a “human-in-the-loop.”

In Block 611, the gas is injected into the well using the gas-lift system according to the optimal set of gas-lift parameters. The step may be performed automatically at the conclusion of the step described in Block 609 or may be initiated manually.

Embodiments of the present disclosure may provide at least one of the following advantages. As noted, complex interactions between the components and sub-components of a well site with gas-lift, above-ground installations (e.g., gas pump), below-ground installations and structures (e.g., well, reservoir) environmental and physical conditions (e.g., seismic activity, or weather affecting the time or resources available to an offshore well, such as available gas for injection) exist and affect the processes of the well site with gas-lift. As such, configuring a gas-lift system (including, in one or more embodiments, a gas pump) for automatic, optimal gas-lifting at a well site with gas-lift is a difficult and laborious task. The gas-lifting objective may be further made difficult in offshore operations which require securing a barge and exacerbating the effects of weather, providing a narrow interval of time to conduct the gas-lifting. Further, the state and behavior of well site with gas-lift may be transient according to activity within the well including injection but also fluid extraction, seismic activity affecting the reservoir, and changing weather conditions generally, all of which may or may not vary depending on the location and history of a given well. Gas-lift stability, in particular, is challenging to maintain. Gas-lift stability refers to the ability of a gas-lift system to maintain consistent and efficient operations over time without experiencing issues that could disrupt production. Instabilities in gas-lift operations can lead to reduced production rates, increased operational costs, and potential damage to equipment or personnel.

A notable example of gas-lift instability is a slug flow, in which large pockets of gas and liquid alternate in the production tubing, causing rapid changes in pressure and flow rates. A slug flow may lead to inefficient lifting, decreased production and wear and tear on well equipment. The gas injection rate is typically balanced in order to create sufficient buoyancy to lift the fluids without causing excessive gas breakout at the surface. Over-injecting gas might lead to inefficient gas usage, while under-injecting gas could result in poor production rates. The gas-to-liquid ratio may also play a significant role in gas-lifted hydrocarbon production, as injecting too much gas may lead to inefficient fluid separation at the surface, while injecting too little gas may result in poor lifting performance. By continuously receiving and processing gas-lift data with a ML model, the well site with gas-lift 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 well site with gas-lift and the adjoining facilities. This, in turn, makes optimal use of the gas that is available for injection in gas-lifting procedures, further saving cost. In addition, since the adjustment of the set of gas-lift parameters may be carried out automatically without direct human involvement, safety at well site with gas-lifts, both onshore and offshore, is improved as well. The probability of success of the gas-lifting operation is further enhanced by embodiments of the instant disclosure, in that multiple predictions for the optimal gas-lift parameters are provided. Lastly, the output of the ML model and optimizer may be digitally formatted in a way that is readily accessible and useful to professionals in the oil and gas industry (e.g., in a Microsoft Excel® “Macro Visual Basic Application” (VBA) and in the software PROSPER).

Embodiments may be implemented on a computer system. FIG. 7 is a block diagram of a computer system (702) 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 (702) 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 ML model may be implemented on an edge computing device in order to quickly provide optimal sets of gas-lift parameters to associated devices or their controllers.

Additionally, the computer (702) 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 (702), including digital data, visual, or audio information (or a combination of information), or a GUI.

The computer (702) 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 (702) 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 (702) 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 (702) 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 (702) can receive requests over network (730) from a client application (for example, executing on another computer (702) 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 (702) 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 (702) can communicate using a system bus (703). In some implementations, any or all of the components of the computer (702), both hardware or software (or a combination of hardware and software), may interface with each other or the interface (704) (or a combination of both) over the system bus (703) using an application programming interface (API) (712) or a service layer (713) (or a combination of the API (712) and service layer (713). The API (712) may include specifications for routines, data structures, and object classes. The API (712) 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 (713) provides software services to the computer (702) or other components (whether or not illustrated) that are communicably coupled to the computer (702). The functionality of the computer (702) may be accessible for all service consumers using this service layer. Software services, such as those provided by the service layer (713), 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 (702), alternative implementations may illustrate the API (712) or the service layer (713) as stand-alone components in relation to other components of the computer (702) or other components (whether or not illustrated) that are communicably coupled to the computer (702). Moreover, any or all parts of the API (712) or the service layer (713) 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 (702) includes an interface (704). Although illustrated as a single interface (704) in FIG. 7, two or more interfaces (704) may be used according to particular needs, desires, or particular implementations of the computer (702). The interface (704) is used by the computer (702) for communicating with other systems in a distributed environment that are connected to the network (730). Generally, the interface (704) includes logic encoded in software or hardware (or a combination of software and hardware) and operable to communicate with the network (730). More specifically, the interface (704) may include software supporting one or more communication protocols associated with communications such that the network (730) or interface's hardware is operable to communicate physical signals within and outside of the illustrated computer (702).

The computer (702) includes at least one computer processor (705). Although illustrated as a single computer processor (705) in FIG. 7, two or more processors may be used according to particular needs, desires, or particular implementations of the computer (702). Generally, the computer processor (705) executes instructions and manipulates data to perform the operations of the computer (702) and any algorithms, methods, functions, processes, flows, and procedures as described in the instant disclosure.

The computer (702) also includes a memory (706) that holds data for the computer (702) or other components (or a combination of both) that can be connected to the network (730). The memory may be a non-transitory computer-readable medium. For example, memory (706) can be a database storing data consistent with this disclosure. Although illustrated as a single memory (706) in FIG. 7, two or more memories may be used according to particular needs, desires, or particular implementations of the computer (702) and the described functionality. While memory (706) is illustrated as an integral component of the computer (702), in alternative implementations, memory (706) can be external to the computer (702).

The application (707) is an algorithmic software engine providing functionality according to particular needs, desires, or particular implementations of the computer (702), particularly with respect to functionality described in this disclosure. For example, application (707) can serve as one or more components, modules, applications, etc. Further, although illustrated as a single application (707), the application (707) may be implemented as multiple applications (707) on the computer (702). In addition, although illustrated as integral to the computer (702), in alternative implementations, the application (707) can be external to the computer (702).

There may be any number of computers (702) associated with, or external to, a computer system containing computer (702), wherein each computer (702) communicates over network (730). 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 (702), or that one user may use multiple computers (702).

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.

Claims

1. A method, comprising: obtaining gas-lift data from a well site with gas-lift, the well site with gas-lift comprising an oil and gas well with access to a hydrocarbon reservoir and a gas-lift system comprising a gas pump;obtaining a set of gas-lift parameters related to the well site with gas-lift;determining, with a machine learning (ML) model, a predicted gas-lift based on the gas-lift data and the set of gas-lift parameters;determining, with an optimizer applied to the ML model, an optimal set of gas-lift parameters such that the predicted gas-lift is optimized;adjusting, automatically, the set of gas-lift parameters to the optimal set of gas-lift parameters; andinjecting gas into the well using the gas-lift system according to the optimal set of gas-lift parameters.

2. The method of claim 1, wherein the gas-lift data comprises well data describing a gas-to-oil ratio inside the well and a wellhead pressure.

3. The method of claim 1, wherein the set of gas-lift parameters comprise gas injection parameters, the gas injection parameters comprising a rate of gas injection and a depth of gas injection.

4. The method of claim 3, wherein the gas injection parameters comprise defining a fluid conduit medium.

5. The method of claim 1, wherein the ML model is trained using empirical well data, the empirical well data comprising a plurality of histories of well behavior.

6. The method of claim 1, wherein the ML model is trained using simulated well data.

7. The method of claim 1, wherein the ML model uses a random forest algorithm.

8. A system, comprising: a well site with gas-lift comprising an oil and gas well with access to a hydrocarbon reservoir and a gas-lift system comprising a gas pump;a plurality of field devices disposed throughout the well site with gas-lift, the plurality of field devices gathering gas-lift data from the well site with gas-lift;a control system configured to adjust one or more field devices in the plurality of field devices and the gas-lift system; anda computer configured to: obtain the gas-lift data from the well and from the reservoir,obtain a set of gas-lift parameters for the well site with gas-lift,determine, with a machine learning (ML) model, a predicted gas-lift based on the gas-lift data and gas-lift parameters,determine, with an optimizer applied to the ML model, an optimal set of gas-lift parameters such that the predicted gas-lift is optimized,adjust, automatically, the set of gas-lift parameters to the optimal set of gas-lift parameters, andtransmit a signal to inject gas into the well using the gas-lift system according to the optimal set of gas-lift parameters.

9. The system of claim 8, wherein the gas-lift data comprises well data describing a gas-to-oil ratio inside the well and a wellhead pressure.

10. The system of claim 8, wherein the set of gas-lift parameters comprise gas injection parameters, the gas injection parameters comprising a rate of gas injection and a depth of nitrogen injection.

11. The system of claim 10, wherein the gas injection parameters comprise defining a fluid conduit medium.

12. The system of claim 8, wherein the ML model is trained using empirical well data, the empirical well data comprising a plurality of histories of well behavior.

13. The system of claim 8, wherein the ML model is trained using simulated well data.

14. A non-transitory computer-readable memory comprising computer-executable instructions stored thereon that, when executed on a processor, cause the processor to perform steps comprising: obtaining gas-lift data from a well site with gas-lift, the well site with gas-lift comprising an oil and gas well with access to a hydrocarbon reservoir and a gas-lift system comprising a gas pump,obtaining a set of gas-lift parameters related to the well site with gas-lift,determining, with a machine learning (ML) model, a predicted gas-lift based on the gas-lift data and the set of gas-lift parameters,determining, with an optimizer applied to the ML model, an optimal set of gas-lift parameters such that the predicted gas-lift is optimized,adjusting, automatically, the set of gas-lift parameters to the optimal set of gas-lift parameters, andtransmitting a signal to inject gas into the well using the gas-lift system according to the optimal set of gas-lift parameters.

15. The non-transitory computer-readable memory of claim 14, wherein the gas-lift data comprises well data describing a gas-to-oil ratio inside the well and a wellhead pressure.

16. The non-transitory computer-readable memory of claim 14, wherein the set of gas-lift parameters comprise gas injection parameters, the gas injection parameters comprising a rate of gas injection and a depth of gas injection.

17. The non-transitory computer-readable memory of claim 16, wherein the gas injection parameters comprise defining a fluid conduit medium.

18. The non-transitory computer-readable memory of claim 14, wherein the ML model is trained using empirical well data, the empirical well data comprising a plurality of histories of well behavior.

19. The non-transitory computer-readable memory of claim 14, wherein the ML model is trained using simulated well data.

20. The non-transitory computer-readable memory of claim 14, wherein the ML model uses a random forest algorithm.