Real-time inference and uncertainty quantification of friction for coiled tubing operations

Abstract

Certain embodiments of the present disclosure include a method that includes receiving data relating to a coiled tubing operation in substantially real time during the coiled tubing operation. The method also includes iteratively generating an estimated friction coefficient between coiled tubing and a wellbore within which the coiled tubing is disposed during the coiled tubing operation based at least in part on the data relating to the coiled tubing operation. The method further includes estimating one or more parameters of the coiled tubing operation during the coiled tubing operation based at least in part on the estimated friction coefficient.

Description

REAL-TIME INFERENCE AND UNCERTAINTY QUANTIFICATION OF FRICTION FOR COILED TUBING OPERATIONSCROSS-REFERENCE TO RELATED APPLICATION

[0001] This application claims priority to and the benefit of U.S. Provisional Patent Application Serial No. 63 / 605,752, entitled “Real-Time Inference and Uncertainty Quantification of Friction for Coiled Tubing Operations,” filed December 4, 2023, which is hereby incorporated by reference in its entirety for all purposes.BACKGROUND

[0002] The present disclosure generally relates to techniques for estimating friction for coiled tubing operations.

[0003] This section is intended to introduce the reader to various aspects of art that may be related to various aspects of the present techniques, which are described and / or claimed below. This discussion is believed to be helpful in providing the reader with background information to facilitate a better understanding of the various aspects of the present disclosure. Accordingly, it should be understood that these statements are to be read in this light, and not as an admission of any kind.

[0004] Tubing force models (TFMs) or tubing force analysis (TFA) models are prediction models that have been used to predict risks involved in coiled tubing (CT) operations, which is useful for reducing operational risk and tracking the remaining useful life (RUL) of CT.However, several model parameters, such as friction coefficient and surface stripper friction, are difficult to obtain and are prone to human errors, which may lead to inaccurate predictions. Therefore, these model parameters must be input manually during or after the operation, whichcontributes to increased cost due to the requirement of additional manpower and time.Therefore, a method to automate the process for inferring and quantifying such model parameters for coiled tubing operations would be useful to alleviate the above concerns.SUMMARY

[0005] A summary of certain embodiments described herein is set forth below. It should be understood that these aspects are presented merely to provide the reader with a brief summary of these certain embodiments and that these aspects are not intended to limit the scope of this disclosure.

[0006] Certain embodiments of the present disclosure include a method that includes receiving data relating to a coiled tubing operation in substantially real time during the coiled tubing operation. The method also includes iteratively generating an estimated friction coefficient between coiled tubing and a wellbore within which the coiled tubing is disposed during the coiled tubing operation based at least in part on the data relating to the coiled tubing operation. The method further includes estimating one or more parameters of the coiled tubing operation during the coiled tubing operation based at least in part on the estimated friction coefficient.

[0007] Certain embodiments of the present disclosure also include a processing and control system that includes one or more processors configured to execute processor-executable instructions. The processor-executable instructions, when executed by the one or more processors, cause the processing and control system to receive data relating to a coiled tubing operation in substantially real time during the coiled tubing operation, to iteratively generate an estimated friction coefficient between coiled tubing and a wellbore within which the coiledtubing is disposed during the coiled tubing operation based at least in part on the data relating to the coiled tubing operation, and to estimate one or more parameters of the coiled tubing operation during the coiled tubing operation based at least in part on the estimated friction coefficient.

[0008] Certain embodiments of the present disclosure also include a method that includes creating a data set that includes a ground truth friction value between coiled tubing and a wellbore within which the coiled tubing is disposed, and surface weight exerted on the coiled tubing. The method also includes inputting the data set into a nonlinear estimation model. The method may further includes inputting a randomized friction coefficient into the nonlinear estimation model. In addition, the method may include utilizing the nonlinear estimation model to iteratively generate an estimated friction coefficient by adjusting the randomized friction coefficient based on the data set.

[0009] Various refinements of the features noted above may be undertaken in relation to various aspects of the present disclosure. Further features may also be incorporated in these various aspects as well. These refinements and additional features may exist individually or in any combination. For instance, various features discussed below in relation to one or more of the illustrated embodiments may be incorporated into any of the above-described aspects of the present disclosure alone or in any combination. The brief summary presented above is intended to familiarize the reader with certain aspects and contexts of embodiments of the present disclosure without limitation to the claimed subject matter.BRIEF DESCRIPTION OF THE DRAWINGS

[0010] Various aspects of this disclosure may be better understood upon reading the following detailed description and upon reference to the drawings, in which:

[0011] FIG. 1 illustrates a schematic diagram of an example coiled tubing (CT) system, in accordance with embodiments of the present disclosure;

[0012] FIG. 2 illustrates a well control system including a surface processing system to control the CT system of FIG. 1, in accordance with embodiments of the present disclosure;

[0013] FIG. 3 illustrates a flowchart of a method for inferring a friction coefficient relating to CT operations, in accordance with embodiments of the present disclosure;

[0014] FIG. 4A illustrates measured surface weight versus a physics model-based predicted surface weight as a function of time, in accordance with embodiments of the present disclosure;

[0015] FIG. 4B illustrates an inferred downhole friction over time, in accordance with embodiments of the present disclosure;

[0016] FIG. 5A illustrates measured surface weight, a data-based predicted weight, and data uncertainty bounds as a function of time, in accordance with embodiments of the present disclosure;

[0017] FIG. 5B illustrates a conveyance speed of CT pipe and a data-driven anomaly alarm, in accordance with embodiments of the present disclosure;

[0018] FIG. 5C illustrates measured surface weight against depth, in accordance with embodiments of the present disclosure; and

[0019] FIG. 6 illustrates a flow diagram of a method of estimating friction for CT operations, in accordance with embodiments of the present disclosure.DETAILED DESCRIPTION

[0020] One or more specific embodiments of the present disclosure will be described below. These described embodiments are only examples of the presently disclosed techniques.Additionally, in an effort to provide a concise description of these embodiments, all features of an actual implementation may not be described in the specification. It should be appreciated that in the development of any such actual implementation, as in any engineering or design project, numerous implementation-specific decisions must be made to achieve the developers’ specific goals, such as compliance with system -related and business-related constraints, which may vary from one implementation to another. Moreover, it should be appreciated that such a development effort might be complex and time consuming, but would nevertheless be a routine undertaking of design, fabrication, and manufacture for those of ordinary skill having the benefit of this disclosure.

[0021] When introducing elements of various embodiments of the present disclosure, the articles “a,” “an,” and “the” are intended to mean that there are one or more of the elements. The terms “comprising,” “including,” and “having” are intended to be inclusive and mean that there may be additional elements other than the listed elements. Additionally, it should be understood that references to “one embodiment” or “an embodiment” of the present disclosure are not intended to be interpreted as excluding the existence of additional embodiments that also incorporate the recited features.

[0022] As used herein, the terms “connect,” “connection,” “connected,” “in connection with,” and “connecting” are used to mean “in direct connection with” or “in connection with via one or more elements”; and the term “set” is used to mean “one element” or “more than one element.” Further, the terms “couple,” “coupling,” “coupled,” “coupled together,” and “coupledwith” are used to mean “directly coupled together” or “coupled together via one or more elements.” As used herein, the terms “up” and “down,” “uphole” and “downhole”, “upper” and“lower,” “top” and “bottom,” and other like terms indicating relative positions to a given point or element are utilized to more clearly describe some elements. Commonly, these terms relate to a reference point as the surface from which drilling operations are initiated as being the top (e.g., uphole or upper) point and the total depth along the drilling axis being the lowest (e.g., downhole or lower) point, whether the well (e.g., wellbore, borehole) is vertical, horizontal or slanted relative to the surface. In addition, the term “interval” with respect to coiled tubing (CT) pipe is used to mean a particular axial portion along an axial length of the CT pipe. In addition, the term “a priori data” is used to mean data that is determined based on theoretical deduction rather than empirical measurement.

[0023] In addition, as used herein, the terms “real time”, ’’real-time”, or “substantially real time” may be used interchangeably and are intended to described operations (e.g., computing operations) that are performed without any human-perceivable interruption between operations. For example, as used herein, data relating to the systems described herein may be collected, transmitted, and / or used in control computations in “substantially real time” such that data readings, data transfers, and / or data processing steps occur once every second, once every 0.1 second, once every 0.01 second, or even more frequent, during operations of the systems (e.g., while the systems are operating). In addition, as used herein, the terms “automatic”, “automatically”, and “automated” are intended to describe operations that are performed or caused to be performed, for example, by a processing / control system (i.e., solely by the processing / control system, without human intervention). In addition, as used herein, the term“approximately equal to” may be used to mean values that are relatively close to each other (e ., within 5%, within 2%, within 1%, within 0.5 %, or even closer, of each other).

[0024] As described above, coiled tubing (CT) is used in the intervention of oil and gas wells. CT is often selected as an intervention method for a number of features, including its capacity to pump fluids, its rigidity for delivering extended reach in deviated wells, its pulling and pushing capacity, and its ability to intervene in live wells. For example, CT pipe may be used to convey tools and fluids into a well. CT pipe is a continuous metal tubular with one outer diameter (OD) and one or more internal diameters (IDs) and is manufactured by welding long sections of pipe together. Typically, CT pipe can measure up to 30,000 feet or more in length and have outer diameters of 1.25 inches to 3 or more inches. CT pipe is sufficiently rigid such that it can be pushed into a well and sufficiently strong to withstand relatively high differential pressures in the flowpath of the CT pipe and external to the CT pipe. Additionally, CT pipe is sufficiently flexible that it can be spooled on a CT reel and fed through a gooseneck.

[0025] For many years, tubing force models (TFMs) have been used to de-risk CT operations and track the CT pipe life. However, several model parameters (e.g., the friction coefficient and the surface stripper friction) are relatively difficult to obtain. Incorrect input parameters can lead to inaccurate TFM predictions. Therefore, these parameters generally have to be fitted manually during or after the CT jobs.

[0026] In contrast, the embodiments described herein are capable of automatically inferring such parameters during a CT job using real-time data. Certain embodiments described herein utilize an Unscented Kalman filter (UKF). In certain embodiments, the TFM may predict the surface weight from input parameters. Then, the UKF may solve the “inverse problem” byobserving the relatively noisy surface data and inferring the unknown parameters automatically in substantially real-time.

[0027] In addition, in certain embodiments, the techniques descried herein may also rigorously compute the uncertainty bounds of the inferred parameters and the predicted surface weight. This enables autonomous anomaly detection such as stuck pipe events. To test the inference model, in certain embodiments, a synthetic data set with well-defined ground truth friction value and analytic surface weight may be created first. The noise-corrupted synthetic data may be fed into the customized UKF framework with an incorrect initial guess of the friction. The UKF framework may incrementally and correctly adjust the friction to the groundtruth friction value in a few hundred iterations at 1 Hz.

[0028] Next, actual data from previous CT jobs may be iteratively replayed into the UKF framework. No ground truth friction value existed for these datasets. However, with the realtime inferred friction, the model was observed to provide a much better surface weight prediction as compared to legacy models without parameter inference.

[0029] As such, the embodiments described herein enable inference of unknown parameters automatically and continuously in substantially real-time during CT operations. UKF provides clear advantages as compared to legacy manual, tedious and subjective fitting-based techniques. In addition, the techniques described herein enable autonomous anomaly detection and CT edge automation. In addition, the techniques described herein leverage real-time noisy data to substantially improve the prediction power of legacy physics models. The techniques described herein also have great potential for application in other types of conveyance operations in the future.

[0030] With the foregoing in mind, FIG. 1 illustrates a schematic diagram of an example CT system 10. As illustrated, in certain embodiments, a CT string 12 may be run into a wellbore 14 that traverses a hydrocarbon-bearing formation 16 (i.e., reservoir). While certain elements of theCT system 10 are illustrated in FIG. 1, other elements of the CT system 10 (e.g., blow-out preventers, wellhead “tree”, etc.) may be omitted for clarity of illustration. In certain embodiments, the CT system 10 includes an interconnection of pipes, including vertical and / or horizontal casings, CT pipe 20, and so forth, that connect to a surface facility 22 at the surface 24 of the CT system 10. In certain embodiments, the CT pipe 20 extends inside drill pipe, tubing, casing, or liner (collectively referred to as a tubular 18) and terminates at a tubing head (not shown) at or near the surface 24. In addition, in certain embodiments, the tubular 18 contacts the wellbore 14 and terminates at a casing head (not shown) at or near the surface 24.

[0031] In certain embodiments, a bottom hole assembly (“BHA”) 26 may be run inside the tubular 18 by the CT pipe 20. As illustrated in FIG. 1, in certain embodiments, the BHA 26 may include a downhole motor 28 that operates to rotate a drill bit 30 (e.g., during drilling operations) or other downhole tools. In certain embodiments, the downhole motor 28 may be driven by hydraulic forces carried in fluid supplied from the surface 24 of the CT system 10. In certain embodiments, the BHA 26 may be connected to the CT pipe 20, which is used to run the BHA 26 to a desired location within the wellbore 14. It is also contemplated that, in certain embodiments, the rotary motion of the drill bit 30 may be driven by rotation of the CT pipe 20 effectuated by a rotary table or other surface-located rotary actuator. In such embodiments, the downhole motor 28 may be omitted.

[0032] In certain embodiments, the CT pipe 20 may also be used to deliver fluid 32 to the drill bit 30 through an interior of the CT pipe 20 to aid in the drilling process and carry cuttingsand possibly other fluid or solid components in return fluid 34 that flows up the annulus between the CT pipe 20 and the tubular 18 (or via a return flow path provided by the CT pipe 20, in certain embodiments) for return to the surface facility 22. It is also contemplated that the return fluid 34 may include remnant proppant (e.g., sand) or possibly rock fragments that result from a hydraulic fracturing application, and flow within the CT system 10. Under certain conditions, fracturing fluid and possibly hydrocarbons (oil and / or gas), proppants and possibly rock fragments may flow from the fractured formation 16 through perforations in a newly opened interval and back to the surface 24 of the CT system 10 as part of the return fluid 34. In certain embodiments, the BHA 26 may be supplemented behind the rotary drill by an isolation device such as, for example, an inflatable packer that may be activated to isolate the zone below or above it and enable local pressure tests. In addition, in certain embodiments, the BHA 26 may include a tractor or agitating system that is capable of improving reach and WOB of the BHA 26 during CT operations.

[0033] As such, in certain embodiments, the CT system 10 may include a downhole well tool 36 that is moved along the wellbore 14 via the CT pipe 20. In certain embodiments, the downhole well tool 36 may include a variety of drilling / cutting tools coupled with the CT pipe 20. In the illustrated embodiment, the downhole well tool 36 includes the drill bit 30, which may be powered by the downhole motor 28 (e.g., a positive displacement motor (PDM), or other hydraulic motor) of the BHA 26. In certain embodiments, the wellbore 14 may be an openhole wellbore or a cased wellbore defined by the tubular 18. In addition, in certain embodiments, the wellbore 14 may be vertical or horizontal or inclined. It should be noted the downhole well tool 36 may be part of various types of BHAs 26 coupled to the CT pipe 20.

[0034] Despite recent advancements in data science, edge computing, and automation technologies, conveying a CT string 12 to a target depth inside a wellbore 14 remains a largely manual process. The primary challenge in automation lies in the need to monitor and respond promptly to downhole anomalies that can disrupt the system in seconds. For example, when CT pipe 20 becomes stuck, operators may only have a fraction of a second to activate the brakes to halt downhole movement of the CT string 12. Any delay is likely to result in cable or tubing breakage, leading to significant downtime. However, short reaction time is not the only hurdle towards automation. In most CT jobs, operators sometimes lack real-time downhole sensors for monitoring. For example, they often may only have a relatively noisy cable tension sensor on the surface (or a surface load cell for coiled tubing operations), which could be kilometers away from the downhole anomalies. Operators must decipher these noisy surface signals, heavily relying on their experience to identify signs of a downhole anomaly. To free the operators from this demanding task, the embodiments described herein present robust and autonomous anomaly detection algorithms. These algorithms are key enablers for a safe automated conveyance job.

[0035] As also illustrated in FIG. 1, in certain embodiments, to aid this autonomous anomaly detection, the CT system 10 may include a downhole sensor package 38 having multiple downhole sensors 40. In certain embodiments, the sensor package 38 may be mounted along the CT string 12, although certain downhole sensors 40 may be positioned at other downhole locations in other embodiments. In addition, in certain embodiments, downhole sensors 40 disposed on the CT pipe 20 may be configured to detect downhole flow rates, downhole temperatures, and downhole pressures, and so forth, in the wellbore 14. In addition, in certain embodiments, downhole sensors 40 disposed on the tubular 18 may be configured to detectdownhole temperatures, downhole pressures, axial load (or “weight”) and torque applied on the bit, casing collar locators (CCLs), resistivity, and so forth, in the wellbore 14.

[0036] In certain embodiments, data from the downhole sensors 40 may be relayed uphole to a processing / control system 42 (e.g., a computer-based processing system) disposed at the surface 24 and / or other suitable location of the CT system 10. In certain embodiments, the data may be relayed uphole in substantially real time (e.g., relayed while it is detected by the downhole sensors 40 during operation of the downhole well tool 36) via a wired or wireless telemetric control line 44, and this real-time data may be referred to as edge data. In certain embodiments, the telemetric control line 44 may be in the form of an electrical line, fiber-optic line, or other suitable control line for transmitting data signals. In certain embodiments, the telemetric control line 44 may be routed along an interior of the CT pipe 20, within a wall of the CT pipe 20, or along an exterior of the CT pipe 20. In addition, as described in greater detail herein, additional data (e.g., surface data) may be supplied by surface sensors 46 and / or stored in a memory location 48. By way of example, historical data and other useful data may be stored in the memory location 48 such as a cloud storage 50.

[0037] As illustrated, in certain embodiments, the CT pipe 20 may be deployed from a CT reel 55 of a CT unit 52 and delivered downhole via an injector head 54. In certain embodiments, the injector head 54 may be controlled to slack off or pick up the CT pipe 20 so as to control the tubing string weight and, thus, the weight-on-bit (WOB) acting on the drill bit 30 (or the downhole well tool 36). In certain embodiments, the downhole well tool 36 may be moved along the wellbore 14 via the CT pipe 20 under control of the injector head 54 so as to apply a desired tubing weight and, thus, to achieve a desired rate of penetration (ROP) as the drill bit 30 is operated. Depending on the specifics of a given application, various types of data may becollected downhole, and transmitted to the processing / control system 42 in substantially real time to facilitate improved operation of the downhole well tool 36. For example, as described in greater detail herein, the data may be used to fully or partially automate downhole operations, to optimize the downhole operations, and / or to provide more accurate predictions regarding components or aspects of the downhole operations.

[0038] In certain embodiments, fluid 32 may be delivered downhole under pressure from a pump unit 56. In certain embodiments, the fluid 32 may be delivered by the pump unit 56 through the downhole motor 28 to power the downhole motor 28 and, thus, the drill bit 30. In certain embodiments, the return fluid 34 is returned uphole, and this flow back of the return fluid 34 is controlled by suitable flowback equipment 58. In certain embodiments, the flowback equipment 58 may include chokes and other components / equipment used to control flow back of the return fluid 34 in a variety of applications, including well treatment applications.

[0039] As described in greater detail herein, the CT unit 52, the injector head 54, the pump unit 56, and the flowback equipment 58 may include advanced surface sensors 46, actuators, and local controllers, such as PLCs, which may cooperate together to provide sensor data to receive control signals from, and generate local control signals based on communications with, respectively, the processing / control system 42. In certain embodiments, as described in greater detail herein, the surface sensors 46 may include flow rate, pressure, and fluid rheology sensors 46, among other types of sensors. In addition, as described in greater detail herein, the actuators may include actuators for pump and choke control of the pump unit 56 and the flowback equipment 58, respectively, among other types of actuators.

[0040] In certain embodiments, surface sensors 46 of the CT unit 52 may be configured to detect positions of the CT pipe 20, weights of the CT pipe 20, and so forth. In addition, incertain embodiments, surface sensors 46 of the injector head 54 may be configured to detect wellhead pressure, and so forth. In addition, in certain embodiments, surface sensors 46 of the pump unit 56 may be configured to detect pump pressures, pump flow rates, and so forth. In addition, in certain embodiments, surface sensors 46 of the flowback equipment 58 may be configured to detect fluids production rates, solids production rates, and so forth.

[0041] The autonomous anomaly detection algorithms described herein may include two main components. The first is generating uncertainty-aware predictions for measurable variables. For example, this could be the surface cable tension or tubing weight predictions and their expected normal error ranges. The second is comparing the real-time measurements to the predictions and their uncertainty bounds. In a simple implementation, any measurement falling outside the bounds may be considered an anomaly.

[0042] For the first component, the embodiments described herein use two approaches to generate the uncertainty-aware prediction:1. Data-based prediction. A conveyance job can last for hours to days. At any time during the job, previous measurements may be checked either based on time or on depth to make the prediction. For example, by analyzing the trend and the noise of the measured surface tension of the past 30 minutes or 100 meters, the expected surface tension at the current time may be estimated. This approach uses acquired data to classify normal data versus outliers. While less interpretable compared to a physics model-based prediction, this does not require model calibrations and is less susceptible to errors arising from any simplifying assumptions in a physics model. Moreover, there is no need to restrict analysis to data of the current job. Rather, the algorithms described herein also check thedata of past jobs in the same well or similar nearby wells to refine the prediction uncertainties.2. Physics model-based prediction. For years, surface tension / weight models (i.e., that model surface weight and / or surface tension) have been used for conveyance planning. These models use force balance principles to predict the surface tension / weight as a function of depth of the CT string 12. The challenges of adapting these legacy models for anomaly detection are the following: (1) They require real-time parameter calibrations. For example, the friction coefficient along the well is typically unknown and may vary along the well depth. (2) There is little prior work to quantify the uncertainty of these models. The uncertainties can come from parameter uncertainty or model assumptions and fidelities; both of which nonlinearly affect the surface tension / weight predictions. Despite the challenges, there are advantages to this approach:(1) The algorithm does not need to wait for data collection before making a prediction.(2) It is more interpretable. For example, a higher tension during pulling out of hole (POOH) leads to a higher well friction. Automation can use this insight to take actions.

[0043] FIG. 2 illustrates a well control system 60 that may include the processing / control system 42 to control the CT system 10 described herein, for example, to provide the autonomous anomaly detection algorithms described in greater detail herein. In certain embodiments, the processing / control system 42 may include one or more analysis modules 62 (e.g., a program of computer-executable instructions and associated data) that may be configured to perform various functions of the embodiments described herein. In certain embodiments, to perform these various functions, the one or more analysis modules 62 may execute on one or more processors 64 of the processing / control system 42, which may be connected to one or more storage media66 of the processing / control system 42. Indeed, in certain embodiments, the one or more analysis modules 62 may be stored in the one or more storage media 66.

[0044] In certain embodiments, the computer-executable instructions of the one or more analysis modules 62, when executed by the one or more processors 64, may cause the one or more processors 64 to generate one or more models (e.g., including the FM described in greater detail herein). Such models may be used by the processing / control system 42 to predict values of operational parameters that may or may not be measured (e.g., using gauges, sensors) during CT operations.

[0045] In certain embodiments, the one or more processors 64 may include a microprocessor, a microcontroller, a processor module or subsystem, a programmable integrated circuit, a programmable gate array, a digital signal processor (DSP), or another control or computing device. In certain embodiments, the one or more processors 64 may include machine learning and / or artificial intelligence (Al) based processors. In certain embodiments, the one or more storage media 66 may be implemented as one or more non-transitory computer-readable or machine-readable storage media. In certain embodiments, the one or more storage media 66 may include one or more different forms of memory including semiconductor memory devices such as dynamic or static random access memories (DRAMs or SRAMs), erasable and programmable read-only memories (EPROMs), electrically erasable and programmable read-only memories (EEPROMs) and flash memories; magnetic disks such as fixed, floppy and removable disks; other magnetic media including tape; optical media such as compact disks (CDs) or digital video disks (DVDs); or other types of storage devices. Note that the computer-executable instructions and associated data of the analysis module(s) 62 may be provided on one computer-readable or machine-readable storage medium of the storage media 66, or alternatively, may be provided onmultiple computer-readable or machine-readable storage media distributed in a large system having possibly plural nodes. Such computer-readable or machine-readable storage medium or media are considered to be part of an article (or article of manufacture), which may refer to any manufactured single component or multiple components. In certain embodiments, the one or more storage media 66 may be located either in the machine running the machine-readable instructions, or may be located at a remote site from which machine-readable instructions may be downloaded over a network for execution.

[0046] In certain embodiments, the processor(s) 64 may be connected to a network interface 68 of the processing / control system 42 to allow the processing / control system 42 to communicate with the multiple downhole sensors 40 and surface sensors 46 described herein, as well as communicate with the actuators 70 and / or PLCs 72 of the surface equipment 74 (e.g., the CT unit 52, the injector head 54, the pump unit 56, the flowback equipment 58, and so forth) and of the downhole equipment 76 (e.g., the BHA 26, the downhole motor 28, the drill bit 30, the downhole well tool 36, and so forth) for the purpose of controlling operation of the CT system 10, as described in greater detail herein. In certain embodiments, the network interface 68 may also facilitate the processing / control system 42 to communicate data to the cloud storage 50 (or other wired and / or wireless communication network) to, for example, archive the data or to enable external computing systems 78 to access the data and / or to remotely interact with the processing / control system 42.

[0047] It should be appreciated that the well control system 60 illustrated in FIG. 2 is only one example of a well control system, and that the well control system 60 may have more or fewer components than shown, may combine additional components not depicted in the embodiment of FIG. 2, and / or the well control system 60 may have a different configuration orarrangement of the components depicted in FIG. 2. In addition, the various components illustrated in FIG. 2 may be implemented in hardware, software, or a combination of both hardware and software, including one or more signal processing and / or application specific integrated circuits. Furthermore, the operations of the well control system 60 as described herein may be implemented by running one or more functional modules in an information processing apparatus such as application specific chips, such as application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), programmable logic devices (PLDs), systems on a chip (SOCs), or other appropriate devices. These modules, combinations of these modules, and / or their combination with hardware are all included within the scope of the embodiments described herein.

[0048] As described in greater detail herein, the embodiments described herein include methods, systems, and apparatus for real-time inference and uncertainty quantification of friction for CT operations. FIG. 3 illustrates a flowchart of a method 80 for inferring a friction coefficient. As illustrated in FIG. 3, the method 80 may include creating a data set that includes: (1) a ground truth friction value between CT pipe 20 and a wellbore 14, and (2) surface weight exerted on the CT pipe 20 (block 82). The method 80 may also include inputting the data set into a nonlinear estimation model (block 84) and inputting a randomized friction coefficient into the nonlinear estimation model (block 86). Based on the provided data, the nonlinear estimation model may iteratively generate an estimated friction coefficient by adjusting the randomized friction coefficient based on the data set (e.g., including the ground truth friction value and the surface weight exerted on the CT pipe 20) (block 88). For example, in general, the randomized friction coefficient may be brought closer to the ground truth friction value, as described in greater detail herein.

[0049] As described in greater detail herein, in certain embodiments, the nonlinear estimation model may be a customized UKF framework. In addition, in certain embodiments, each iteration of the estimated friction coefficient generated by the nonlinear estimation model may be performed at a frequency of approximately 1 Hz. The accuracy of the nonlinear estimation model may be tested by inputting data sets obtained from preceding CT operations, which do not contain a ground truth friction value. The nonlinear estimation model has been observed to provide a relatively accurate surface weight prediction when compared to legacy models without parameter inference, as described herein.

[0050] In addition, in certain embodiments, two approaches to generate uncertainty aware prediction of friction coefficients, as well as the corresponding predicted surface tension / weight, may be used to extend the techniques described with reference to the method 80 illustrated in FIG. 3, namely, data-based prediction and physics model -based prediction. As described herein, “surface tension / weight” is used to mean surface weight of coiled tubing as, for example, determined based on a tension value on the coiled tubing, as measured by a surface sensor 46.

[0051] The data-based prediction approach will be described now. In certain embodiments, the algorithms may first check if the CT system 10 is running in one of the two steady states: running-in-hole (RIH) or POOH. In certain embodiments, the algorithms may filter out data measurements during transient events because those data exhibit different patterns and cannot be used for predicting the surface tension / weight during steady states. In certain embodiments, the filtered data may be recorded as a triple (as a function of three parameters, for example, tension / weight-depth-time, as the three parameters of tension / weight, depth, and time). In general, depth and time may be used to determine how important the data are in predicting the current tension / weight. Data that were collected either too far away or too close in time or depthmay be assigned a lower importance score for prediction. In certain embodiments, data that are far away are not as good because the data pattern may have changed. In general, data too close are not good because there is a risk of including a currently occurring anomaly that the algorithms have not identified yet. In certain embodiments, the filtered data, with importance scores assigned, may then be input into a robust nonparametric locally estimated scatterplot shooting (LOESS regression) algorithm to make prediction of the current surface tension / weight. The LOESS regression may also simultaneously update the importance score of the past data, which may then be used for the next round of regression.

[0052] In certain embodiments, the error between the LOESS regression and the measurement may be recorded and input into a generalized extreme studentized deviate (GESD) algorithm for outlier detection. GESD algorithms are particularly useful for detecting outliers on a real-time univariate data set that follows an approximately normal distribution. In addition to outlier detection, GESD may simultaneously predict upper and lower bounds of a “normal” error range (e g., one that arises from sensor noise instead of anomalies). These bounds may be used directly to generate the tension / weight prediction bounds.

[0053] The physics model-based prediction approach will be described now. As mentioned herein, physics models that predict the surface tension / weight exist for job planning. The embodiments described herein adapt these physics models, solving two main challenges. The first challenge is the need to continuously calibrate uncertain model parameters (e.g., well friction) and at the same time quantify the uncertainties of the model parameters. A UKF may be used for real-time model parameter inference. The UKF represents the model parameter uncertainty as a multivariate Gaussian distribution and iteratively updates the distribution using noise-characterized surface data. The uncertainty-quantified parameters may then be input intothe physics model to estimate the surface tension / weight and its prediction error ranges. The UKF has been updated with a more intelligent update schedule. The improved version of the UKF tracks new information in the data received since the last inference. Only when enough new information is collected will a new inference be triggered by the UKF. For example, when the system stops, no new information will be contained in the incoming data; therefore, no new inference will be performed.

[0054] The second challenge is to quantify the model prediction uncertainty. Uncertainty comes from: (1) model parameter uncertainty, which can be estimated by using a forward uncertainty estimation model, and (2) model simplification assumptions and fidelity, which can be estimated using the model prediction versus measurement error for the current job and for past jobs with similar features.

[0055] For the final anomaly detection algorithm, the detection from both the data-based prediction approach and the physics model-based prediction approach may be fused, as described in greater detail herein. In addition, error trends may be considered. For example, considerations may include whether the measurement is rapidly and monotonously falling out of the error range and / or whether the measurement is bouncing closely around the prediction bounds. This reduces the false positive alarms.Test Case 1

[0056] In the first example case, the model was run for a coiled tubing cleanout job. The well has a total depth of 5.3 kilometers and its maximum inclination is 96°. FIGS. 4A and 4B illustrate the application of the physics model-based prediction approach to the coiled tubing cleanout job. In particular, FIG. 4A illustrates the measured surface weight 90 (e.g., as measured by surface sensors 46) versus the physics model-based predicted surface weight 92 as a functionof time. The data from time 45 to 60 hours were mostly for the RIH operation with the weight spikes being pull test signals. The rest of the data were the final POOH operation. The spikes correspond to the pull tests during the job. The weight data, received at 1 Hz, were input into the UKF to update the well friction parameters iteratively over time to determine the inferred downhole friction over time, as illustrated in FIG. 4B. The real-time inferred friction was input back into the physics-based weight model to predict the surface weight 92 with the physics model uncertainty bounds 94 used for identifying downhole anomalies, as illustrated in FIG. 4A. Test Case 2

[0057] FIGS. 5A through 5C illustrate the application of the data-based prediction approach and the corresponding alarm for the same coiled tubing cleanout job as in FIGS. 3A and 3B. Instead of showing the entire job, the various plots focus on the first few hours of operation. FIG. 4A illustrates the measured surface weight 90, the data-based predicted surface weight 96, and the data uncertainty bounds 98 as a function of time, similar to the results illustrated in FIG. 3A. It is noted that, unlike the physics model-based prediction, the data-based prediction requires some data collected before the first prediction can be made around 44.7 hours. FIG. 4B illustrates the conveyance speed 100 of the CT pipe 20 (e.g., based on a fixed point of the CT pipe 20) on the left y-axis and the data-driven anomaly alarm 102 on the right y-axis (0 indicating no alarm, -1 indicating a weight drop event, and +1 indicating a weight increase event) as a function of time. It is noted that the data-driven anomaly alarm 102 was triggered for a weight drop event around 45.35 hours (e.g., as illustrated by arrow 104), and it was set to be persistent until later when the weight value returned to normal. This weight drop event can be easily spotted around a depth of 420 meters (e.g., as illustrated by arrow 106) in FIG. 4C, wherethe measured surface weight 90 is plotted against the depth of the CT pipe 20 with the shaded bar 108 showing the corresponding time in hours.

[0058] As such, to automate conveyance, robust algorithms were developed to automatically identify downhole anomalies. Two approaches were developed, the first based on data while the second is based on physics-based tension / weight models. For both prediction approaches, prediction of the surface tension / weight is made together with a quantified uncertainty, which is used to automatically identify anomalies, which may then be used to automatically adjust CT conveyance parameters, such as conveyance speed, and so forth.

[0059] Advantageously, the techniques described herein are capable of inferring data parameters such as the friction coefficient between the CT pipe 20 and the operating environment (i.e., a wellbore 14) in substantially real-time based on surface data. Furthermore, the techniques described herein can also compute the uncertainty bounds of the inferred data parameters and the predicted surface weight. In addition, using the inferred data parameters, anomaly events such as stuck pipe can be autonomously detected. The techniques described herein advantageously enable autonomous anomaly detection as well as CT edge automation, as the techniques leverage real-time data to substantially improve the efficiency and accuracy of TFMs and TFA models.

[0060] FIG. 6 illustrates a flow diagram of a method 110 of estimating friction for CT operations, for example, of the CT system 10 of FIG. 1 by the processing / control system 42 of FIG. 2. As illustrated, in certain embodiments, the method 110 may include receiving data relating to a CT operation (e.g., performed by the CT system 10 of FIG. 1) in substantially real time during the CT operation (block 112). In addition, in certain embodiments, the method 110 may include iteratively generating an estimated friction coefficient between CT pipe 20 of a CTstring 12 and a wellbore 14 within which the CT string 12 is disposed during the CT operation based at least in part on the data relating to the CT operation (block 114). In addition, in certain embodiments, the method 110 may include estimating one or more parameters (e.g., surface tension / weight of the CT string 12, conveyance speed of the CT string 12, and so forth) of the CT operation during the CT operation based at least in part on the estimated friction coefficient (block 116).

[0061] In addition, in certain embodiments, the method 110 may include automatically generating an alarm 102 during the CT operation based at least in part on the one or more estimated parameters. In addition, in certain embodiments, the method 110 may include automatically adjusting one or more operational parameters of the CT operation (e.g., conveyance speed of the CT string 12, in certain embodiments) based at least in part on the one or more estimated parameters. In addition, in certain embodiments, the data relating to the CT operation may include measured tension / weight on the CT string 12, depth of the CT string 12, time during the CT operation, a ground truth friction value, or some combination thereof.

[0062] In addition, in certain embodiments, the method 110 may include utilizing physics model-based algorithms to generate the estimated friction coefficient and to estimate a surface weight of CT string 12 based at least in part on the data relating to the CT operation. In addition, in certain embodiments, the method 110 may include utilizing a customized UKF framework to generate the estimated friction coefficient and to estimate a surface weight of the CT string 12 based at least in part on the data relating to the CT operation. In addition, in certain embodiments, the method 110 may include utilizing data-based algorithms to estimate the one or more parameters of the CT operation based at least in part on the data relating to the CToperation. In addition, in certain embodiments, the method 110 may include iteratively generating uncertainty bounds 94, 98 related to the estimated friction coefficient.

[0063] The specific embodiments described above have been illustrated by way of example, and it should be understood that these embodiments may be susceptible to various modifications and alternative forms. It should be further understood that the claims are not intended to be limited to the particular forms disclosed, but rather to cover all modifications, equivalents, and alternatives falling within the spirit and scope of this disclosure.

[0064] The techniques presented and claimed herein are referenced and applied to material objects and concrete examples of a practical nature that demonstrably improve the present technical field and, as such, are not abstract, intangible or purely theoretical. Further, if any claims appended to the end of this specification contain one or more elements designated as “means for [perform]ing [a function], or “step for [performing [a function], . it is intended that such elements are to be interpreted under 35 U.S.C. § 112(f). However, for any claims containing elements designated in any other manner, it is intended that such elements are not to be interpreted under 35 U.S.C. § 112(f).

Claims

CLAIMS1. A method, comprising: receiving data relating to a coiled tubing operation in substantially real time during the coiled tubing operation; iteratively generating an estimated friction coefficient between coiled tubing and a wellbore within which the coiled tubing is disposed during the coiled tubing operation based at least in part on the data relating to the coiled tubing operation; and estimating one or more parameters of the coiled tubing operation during the coiled tubing operation based at least in part on the estimated friction coefficient.

2. The method of claim 1, comprising automatically generating an alarm during the coiled tubing operation based at least in part on the one or more estimated parameters.

3. The method of claim 1, comprising automatically adjusting one or more operational parameters of the coiled tubing operation based at least in part on the one or more estimated parameters.

4. The method of claim 1, wherein the data relating to the coiled tubing operation comprises measured tension / weight on the coiled tubing, depth of the coiled tubing, time during the coiled tubing operation, or some combination thereof.

5. The method of claim 1, comprising utilizing physics model-based algorithms to generate the estimated friction coefficient and to estimate a surface weight of coiled tubing based at least in part on the data relating to the coiled tubing operation.

6. The method of claim 1, comprising utilizing a customized Unscented Kalman Filter (UKF) framework to generate the estimated friction coefficient and to estimate a surface weight of the coiled tubing based at least in part on the data relating to the coiled tubing operation.

7. The method of claim 1, comprising utilizing data-based algorithms to estimate the one or more parameters of the coiled tubing operation based at least in part on the data relating to the coiled tubing operation.

8. The method of claim 1, comprising iteratively generating uncertainty bounds related to the estimated friction coefficient.

9. A processing and control system, comprising: one or more processors configured to execute processor-executable instructions that, when executed by the one or more processors, cause the processing and control system to: receive data relating to a coiled tubing operation in substantially real time during the coiled tubing operation;iteratively generate an estimated friction coefficient between coiled tubing and a wellbore within which the coiled tubing is disposed during the coiled tubing operation based at least in part on the data relating to the coiled tubing operation; and estimate one or more parameters of the coiled tubing operation during the coiled tubing operation based at least in part on the estimated friction coefficient.

10. The processing and control system of claim 9, wherein the processor-executable instructions, when executed by the one or more processors, cause the processing and control system to automatically generate an alarm during the coiled tubing operation based at least in part on the one or more estimated parameters.

11. The processing and control system of claim 9, wherein the processor-executable instructions, when executed by the one or more processors, cause the processing and control system to automatically adjust one or more operational parameters of the coiled tubing operation based at least in part on the one or more estimated parameters.

12. The processing and control system of claim 9, wherein the data relating to the coiled tubing operation comprises measured tension / weight on the coiled tubing, depth of the coiled tubing, time during the coiled tubing operation, or some combination thereof.

13. The processing and control system of claim 9, wherein the processor-executable instructions, when executed by the one or more processors, cause the processing and control system to utilize physics model-based algorithms to generate the estimated friction coefficientand to estimate a surface weight of coiled tubing based at least in part on the data relating to the coiled tubing operation.

14. The processing and control system of claim 9, wherein the processor-executable instructions, when executed by the one or more processors, cause the processing and control system to utilize a customized Unscented Kalman Filter (UKF) framework to generate the estimated friction coefficient and to estimate a surface weight of the coiled tubing based at least in part on the data relating to the coiled tubing operation.

15. The processing and control system of claim 9, wherein the processor-executable instructions, when executed by the one or more processors, cause the processing and control system to utilize data-based algorithms to estimate the one or more parameters of the coiled tubing operation based at least in part on the data relating to the coiled tubing operation.

16. The processing and control system of claim 9, wherein the processor-executable instructions, when executed by the one or more processors, cause the processing and control system to iteratively generate uncertainty bounds related to the estimated friction coefficient.

17. A method, comprising: creating a data set comprising: a ground truth friction value between coiled tubing and a wellbore within which the coiled tubing is disposed; and surface weight exerted on the coiled tubing;inputting the data set into a nonlinear estimation model; inputting a randomized friction coefficient into the nonlinear estimation model; and utilizing the nonlinear estimation model to iteratively generate an estimated friction coefficient by adjusting the randomized friction coefficient based on the data set.

18. The method of claim 17, comprising automatically adjusting one or more operational parameters of a coiled tubing operation based at least in part on the estimated friction coefficient.

19. The method of claim 17, wherein the nonlinear estimation model is a customized Unscented Kalman Filter (UKF) framework.

20. The method of claim 17, wherein each iteration of the estimated friction coefficient generated by the nonlinear estimation model is performed at a frequency of approximately 1 Hz.

Images