Well Plug Integrity

Some implementations relate to operations for hydraulic fracturing in a well. More specifically, some implementations relate to determining integrity of a plug that may be used in conjunction with operations for hydraulic fracturing in a well.

In oil and gas industry, hydraulic fracturing (“fracking”) is a common method used to increase permeability and thus productivity of the reservoirs. In fracking, a plug may be used in a well as a temporary barrier to isolate a certain well area, to control the flow of the fracturing fluid through perforations of the wellbore casing, and to allow pressurizing of the well area to perform effective hydraulic fracturing. However, plug integrity may be compromised during hydraulic fracturing due to factors such as pressure fluctuations, mechanical stress, and fluid flow. Therefore, plug integrity may be important for success in hydraulic fracturing of the reservoir. The loss of the integrity of the plug (leak-offs, breaks) can cause damage to well components, reduced well productivity, and lose hydraulic treatment stages.

The description that follows may include example systems, methods, techniques, and program flows that embody implementations of the disclosure. However, this disclosure may be practiced without these specific details. For clarity, some well-known instruction instances, protocols, structures, and techniques may not be shown in detail.

In fracking, plugs may be used to fluidically segregate parts of a well.is a sectional view showing a plug deployed in a well during fracking operations. In, a wellincludes a wellbore. Although shown as a horizontal wellbore, the wellboremay be a vertical wellbore or otherwise include one or more sections at any suitable angle. The wellboreincludes active stage perforationsand previous stage perforations. The plugfluidically segregates the active stageof the wellfrom the previous stageof the well. Conventional methods for determining integrity of the plugmay involve logging and testing, tracers, DAS/DTS systems, and other techniques, individually or combined. These conventional techniques may be costly and unavailable during fracking operations.

Some implementations include a learning machine (such as a machine learning model, a machine learning neural network, or other suitable particularized machine) configured to utilize historical data, Distributed Acoustic Sensing (DAS)/Distributed Temperature Sensing (DTS) data, and pressure pulse water hammer analysis data to identify and characterize plug integrity. After the learning machine is trained on the aforementioned data, it may predict plug problems using only water hammer pressure pulse data and without using any conventional methods.

Some implementations include a method for detecting issues with plug integrity (such as leakage) in hydraulic fracturing operations. Specifically, one method may leverage DAS and/or DTS data, historical records, water hammer pressure pulse data, and machine learning (ML) operations to train a learning machine to identify fracking stages that have plug leaks. After training, the trained learning machine may identify a fracking stage that has a leaky plug by estimating the severity of plug leak-off or the presence/absence of plug leaks during hydraulic fracturing processes.

In some implementations, an ML neural network may be trained for a set of water hammer analysis parameters (such as resistivity, decay, amplitude, friction factor, and/or pressure water hammer data itself) along with other pertinent information such as treatment data and wellbore geometrical details. After training, the trained ML neural network may provide a plug integrity indicator. Some implementations may utilize other techniques (other than DAS/DTS) for estimating the plug integrity. For example, some implementations of the trained ML neural network may estimate the plug integrity directly from pressure hammer data analysis parameters (without other information).

By identifying the plug integrity issues, some implementations can prevent lost hydraulic treatment stages, avoid reduction of productivity, and avoid casing damage. Additionally, early detection of plug leakage or failure may enable well operators to react to such plug failure/leakage, such as by changing treatment parameters that may prevent damage and maintain production. Some implementations react to predictions (made by the learning machine) of plug failure by recommending alternative fracking operations that mitigate damage, lost production, or other problems that may arise from the plug failure.

Some implementations may collect data for training a learning machine (such as an ML neural network or other particularized machine). Using fiber optic cable technology, data may be collected in real-time from distributed acoustic sensors (DAS) and distributed temperature sensing (DTS) installed along the wellbore. The DAS sensors and/or DTS may be placed at any suitable location in the well(see also discussion offor more about DAS/DTS).

DAS system may utilize Rayleigh scattering to measure the relative phase of two points along the fiber separated by a distance called the gauge length. Using the relative phase at consecutive time samples, a strain change then may be estimated along the gauge length for the time period along the cable. Therefore, these cables may function as acoustic distributed sensors, recording signals (vibrations) caused by various events (e.g., leaks, pressure changes, mechanical waves/deformations). On the other hand, DTS systems may utilize Raman scattering, especially Anti-Stokes and Stokes components. The Anti-Stokes component may be sensitive while the Stokes component may not be sensitive to the changes in temperature. The ratio of these components may be used to estimate the temperature along the fiber cable in the wellbore.

If a DAS system is installed in a treatment well, DAS data be used to identify whether and where plug leaks occurred: as DAS data may be used to detect fluid flow, detect whether there is a flow past the plug, determine likelihood that the pluglost integrity, and detect leaks in the plug. These results can be recorded and saved to a database or other datastore. In a similar manner, some implementations may use DTS data to identify leaks based on the temperature changes caused by a plug leak. After the flow leaks though the plug, the temperature may be altered and leak detected.

is a graph showing example DAS data that may be used in conjunction with some implementations. The graphshows a DAS “waterfall” plot. The waterfall plot indicates areas of intense fluid flow (dark red color) along the wellbore at various depths (see y-axis of graph) during corresponding time periods (see x-axis of the graph). At approximately 20:55, a new stage treatment started at 3700 m depth. However, as indicated in the graph, fluid may still be flowing in the zone of the previous stage (at approximately 3650 m of depth).

In addition to DAS and/or DTS data, some implementations may use pressure response data. A water hammer is an oscillatory pressure response in a closed system (e.g., pipes and wellbores) caused by an abrupt change in flow rate. A water hammer may be observed in hydraulic fracturing treatments after fluid velocity is rapidly reduced (ramp downs). In hydraulic fracturing, if the fluid flow suddenly changes at the end of the treatment stage, the pressure wave may propagate through the wellbore and interact with the created fracture networks.is a graph showing an example water hammer pressure pulse (psi) in a wellbore. The graphindicates the water hammer pressure pulsefor a welbore casing having a diameter of 4.67 inches at a depth 10341 feet with estimated resistance 20 Pa s/cc.

Some implementations may analyze water hammer characteristics to gain insights into resistance in the well. The resistance (R) is the measure of change of flow with change in pressure

The resistance associated with a water hammer may depend on the system friction (such as friction in the wellbore and hydraulic fracture network) that attenuates the energy of the generated pressure wave. Some implementations may estimate resistance in the well. An estimation of resistance may be based on measurements indicating attenuation of the pressure wave, geometry of the well, diameter of the borehole (such as wellbore), and density of fluid. In this estimation of resistance, the resistance may depend on the integrity of the plug. If the plug's integrity is compromised, the water hammer pressure pulse and estimated resistance may be affected.

In addition to estimating resistance, some implementations may estimate other parameters of the water hammer analysis, such as characterization decay, characterization amplitude, characterization period, Darcey factor. These parameters, water hammer pressure data itself, and other treatment data (well geometry, flow rate, type of fluid, etc.) may be used as inputs to a learning machine configured to make predictions about integrity of one or more plugs in a well.

In some implementations, pressure data may be acquired with 1 Hz or above pressure gauges/sensors in the borehole. The sensor data may be processed, such as by removing noise. Resistivity and other parameters (such as characterization decay, characterization amplitude, characterization period, Darcey factor, etc.) may be estimated from the water hammer pressure pulse. Treatment data may be available and recorded. The treatment data may include, for example, fluid flow rate, properties of pumped fluid, and geometry of the well. DAS/DTS data may be analyzed and plug leaks identified (see DAS/DTS section).

Some implementations utilize learning machines to learn from data patterns and make the decisions without knowledge of the explicit relationships. In some implementations, the learning machine may be configured to learn a function that transforms input data into meaningful predictions or classifications about integrity of plugs in a wellbore. The function may be defined by a neural network, including weights, biases, and activation functions for each neuron.is diagram illustrating an example neural network that may be used in an example learning machine. In, the learning machineincludes the neural network. The neural networkmay include an input layer that intakes information (sometimes referred to as features) about the pressure response, fracking treatments, well geometry, and/or any other suitable information about the well. Although the input layer is shown having four neurons, there may be any suitable number of neurons (hence, any suitable number of features). The neural networkalso may include an output layer that predicts plug integrity based on the information (such as values for the features) that was fed into the input layer.

The neural networkmay perform training based on DAS/DTS data, water hammer data, fracking treatment data, and/or any other suitable data about the well. The process for training the learning machinemay find optimal neural network parameters (such as weights, biases, etc.) that match water hammer analysis data and/or other treatment data with estimated plug integrity determined from DAS/DTS data analysis. Some implementations my use methods other than DAS/DTS for verifying plug integrity (such as well-logging, tracers, and testing that can identify the failure of plug integrity). Failed and successful plug installations may be used for the training.

Operations for training the neural networkmay include creating or obtaining a training data set and inputting the training data set to the neural network. The training data set may include feature samples and prediction samples. The feature samples may include values derived from past pressure pulses in the well (such as values derived from water hammer data analysis), data about hydraulic fracturing operation in the well, well geometry, and/or any other data described herein or otherwise suitable for training the neural network. The feature samples also may include prediction samples including values derived from digital acoustic sensing (DAS) sensors located in the wellbore. During training, the neural networkmay receive a group of the feature samples and make a prediction about plug leakage based on that group of feature samples. The neural networkmay validate or invalidate the prediction based on the prediction sample. For example, the neural networkmay predict no leak but the prediction sample may indicate that a leak was detected. In response to this invalid prediction, the training process may modify the neural network(such as by modifying weights, biases, activation functions, etc.).

In some implementations, predicting plug integrity may be discrete (binary) variable (0—fail, 1—success). In some implementations, predicting plug integrity may be categorized by different scenarios that plug failure can exhibit (leak, break, etc.). In some implementations, predicting plug integrity may be a continuous variable with the degree of severity of plug leak-off.

After the training is complete, the learning machinemay be used to predict plug integrity. For example, the learning machinemay use the water hammer pressure data and analysis parameters to predict the plug integrity, even though DAS data (or other information for estimating plug integrity) is not available.

In some implementations, the learning machinemay be integrated into a computer system.is a block diagram illustrating a computer system that may be utilized with some implementations. In, a computer systemmay include one or more processorsconnected to a system bus. The system busmay be connected to memoryand a network interface. The memorymay include any suitable memory random access memory (RAM), non-volatile memory (e.g., magnetic memory device), and/or any device for storing information and instructions executable by the processor(s). The network interfacemay provide connectivity to any suitable network, such as a wired network, wireless network, satellite network, etc.

The computer systemmay include additional peripheral devices. For example, the computer systemmay include multiple external multiple processors. In some implementations, any of the components can be integrated or subdivided.

The computer systemalso may include a plug evaluator. The plug evaluatormay implement the methods and operations described herein. The plug evaluatormay include a learning machine(as described herein). The learning machinemay include a neural networkor other logic for performing the ML operations described herein. In some implementations, the computer systemmay be included in the well system (such as the well system described with reference toand may cooperate with other components and/or systems to perform the functionality described herein.

The computer systemalso may include a fracking controllerconfigured to perform operations in response to predictions about plug integrity (e.g., see discussion of).

Any component of the computer systemcan be implemented as hardware, firmware, and/or machine-readable media including computer-executable instructions for performing the operations described herein. For example, some implementations include one or more non-transitory machine-readable media including computer-executable instructions including program code configured to perform functionality described herein. Machine-readable media includes any mechanism that provides (e.g., stores and/or transmits) information in a form readable by a machine (e.g., a computer system). For example, tangible machine-readable media includes read only memory (ROM), random access memory (RAM), magnetic disk storage media, optical storage media, flash memory machines, etc. Machine-readable media also includes any media suitable for transmitting software over a network.

Some implementations may perform operations in response to detecting leaky plugs. For example, after the learning machinedetects a plug leak-off, certain actions may be taken to mitigate problems that may arise from the leaky plug. The actions may vary based on the severity of the plug leak-off. If the plug leak-off cannot be controlled, the decision may be to continue to the next treatment on the well. Some actions to seal the plug may involve dropping the diverter, varying the proppant type or proppant concentrations, altering the rate (to increase the pressure to help sealing).describes operations that may be performed in response to detecting a plug leak-off.

is a flow diagram illustrating operations performed in response to detecting a plug leak-off. In, the operations begin at block, where the fracking operation is performed. The fracking operation may include any suitable fracking operation such as injecting, under high pressure, “fracking fluid” into the wellbore.

At block, the pressure pulse (such as the water hammer described herein) is detected.

At block, a prediction is made about plug leak-off severity. For example, the plug evaluatormay make a prediction about leak-off severity (such as no leak, moderate leak, severe leak, etc.) based on data indicating the pressure pulse (such as data collected at block). As noted, the plug evaluator may make predictions about leak-off severity without having access to DAS/DTS data.

At block, a determination is made about whether plug leak-off is acceptable. For example, the plug evaluatormay determine the severity of the leak has not exceeded a severity threshold (i.e., is acceptable) and continue the current fracking plan (such as by continuing operations at block). Otherwise, the plug evaluator may determine plug leak-off severity is unacceptable (has exceeded the severity threshold) and thereby continue operations at block.

At block, a recommendation is made about how to modify operations. For example, the fracking controllermay recommend one or more modifications to the current fracking operation based on the severity of the plug leak-off and based on available resources (such as materials, fluids, and other resources available for fracking operations). Future fracking operations may be modified based on the recommendation made by the control system. Hence, as operations continue at block, fracking operation may be modified per the recommendation.

After performing the flow(repeated any suitable number of times), the efficacy of the control can be confirmed. After sufficient data has been gathered about action of control and its efficacy, the learning machinemay be trained to predict the best control strategy. The learning machinemay consider the current operating conditions (such as available proppant type, available proppant mass to pump, available diverter, volume of fluid remaining to be pumped etc.) and determine an optimal solution under these constraints.

Some implementations may predict plug integrity using a deterministic method. For example, instead of using DAS/DTS data for training the learning machineto make predictions about whether a plug is leaking, some implementations train the learning machineusing the deterministic method (described below).

For example, the plug evaluatormay perform the following computations to determine whether a plug is leaking. The plug evaluatormay use measured resistance to compute the effective number of perforations taking the fracturing fluid. The resistance for each perforation may be calculated by knowing change of rate Q, number of perforations N, discharge coefficient, and hydraulic perforation diameter h=where dis a perforation diameter:

If all hydraulic perforation diameters are the same (h=h), then total resistance becomes

If the plug evaluatorestimated and calculate all parameters with satisfactory precision, we can define a coefficient of plug integrity P, as a square root ratio of estimated and modeled resistances, which gives the ratio of estimated and true number of perforations:

If Pis 1, there is no plug leaks; if Pis less than 1, there is a leak. This measure is not only qualitative, but also quantitative—it represents the portion of the perforations that are taking the fluid: the severity of leak thus is calculated, for example, if P=0.5, only a half of the perforation is taking the fluid (the other half does not).

Theoretically, the plug evaluatorshould not have P>1; that is, the estimated resistance is greater than modeled; otherwise, the parameter estimation as well as resistance is not correctly calculated.

The computer systemmay be part of a larger system for drilling and fracturing well.is an illustration depicting an example multi-well system, according to some implementations. In particular,is a schematic of a multi-well systemthat includes a wellboreand a wellborein a subsurface formation. The wellboreincludes casingand a number of perforationsA-H being made in the casingat different depths to allow reservoir fluids (i.e., oil, water, and gas) from the subsurface formationto flow into the wellbore. Similarly, the wellboreincludes casingand a number of perforationsA-H being made in the casingto allow reservoir fluids (i.e., oil, water, and gas) from the subsurface formationto flow into the wellbore. During hydraulic fracturing operations of the wellbores, fracturing fluid, with or without sand, may be pumped into the subsurface formation, via the perforationsA-H and perforationsA-H, to hydraulically fracture the rock such that reservoir fluid may flow into the wellbore,, respectfully.

In some implementations, one or more sensors may be positioned in a wellbore to obtain measurements while an offset well is being hydraulically fractured. For example, the wellboremay include a fiber optic cableto obtain strain measurements, temperature measurements, derived pressure measurements (from strain measurements), etc. of the subsurface formationwhile the wellboreis being hydraulically fractured. The fiber optic cablemay extend from the wellheadon the surfaceto the subsurface along the wellbore. The fiber optic cablemay be cemented in place in the annular space between the casingof the wellboresand the subsurface formation. The fiber optic cablemay be clamped to the outside of the casingduring deployment and protected by centralizers and cross coupling clamps. The fiber optic cablesmay be included with coiled tubing, wireline, loose fiber using coiled tubing, or gravity deployed fiber coils that unwind the fiber as the coils are moved in the wellbore. The fiber optic cablealso may be deployed with pumped down coils and/or self-propelled containers. Additional deployment options for the fiber optic cablecan include coil tubing and wireline deployed coils where the fiber optic cablesare anchored at the toe of the wellbore. In such implementations the fiber optic cablecan be deployed when the wireline or coiled tubing is removed from the well. The fiber optic cablemay house one or more optical fibers, and the optical fibers may be single mode fibers, multi-mode fibers, or a combination of single mode and multi-mode optical fibers. The distribution of sensors shown inis for example purposes only. Any suitable sensor deployment may be used.

The fiber optic cablemay be used for distributed sensing where acoustic, vibration, strain, and temperature measurements may be collected downhole in the wellbores. The measurements may be collected at various positions distributed along the fiber optic cable. For example, data may be collected every 1-3 ft along the full length of the fiber optic cabledownhole along the horizontal section of the wellbore. Fiber optic interrogation unitof the wellboremay be located on the surfaceof the multi-well system. The fiber optic interrogation unitsmay be directly coupled to the fiber optic cables. Alternatively, the fiber optic interrogation unitsmay be coupled to a fiber stretcher module, wherein the fiber stretcher module is coupled to the fiber optic cable. The fiber optic interrogation unitmay receive measurement values taken and/or transmitted along the length of the fiber optic cablesuch as acoustic, temperature, strain, etc. The fiber optic interrogation unitmay be electrically connected to a digitizer to convert optically transmitted measurements into digitized measurements.

The fiber optic interrogation unitmay operate using various sensing principles including but not limited to amplitude-based sensing systems like DTS, DAS, Low Frequency Distributed Acoustic Sensing (LFDAS), Distributed Vibration Sensing (DVS), and Distributed Strain Sensing (DSS). For example, the DTS system may be based on Raman and/or Brillouin scattering. A DAS system may be a phase sensing-based system based on interferometric sensing using homodyne or heterodyne techniques where the system may sense phase or intensity changes due to constructive or destructive interference. The DAS system may also be based on Rayleigh scattering and in particular coherent Rayleigh scattering. A DSS system may be a strain sensing system using dynamic strain measurements based on interferometric sensors or static strain sensing measurements using Brillouin scattering. DAS systems based on Rayleigh scattering may also be used to detect dynamic strain events. Temperature effects may in some cases be subtracted from both static and/or dynamic strain events, and temperature profiles may be measured using Raman based systems and/or Brillouin based systems capable of differentiating between strain and temperature, and/or any other optical and/or electronic temperature sensors, and/or any other optical and/or electronic temperature sensors, and/or estimated thermal events.

In some implementations, the fiber optic interrogation unitmay measure changes in optical fiber properties between two points in an optical fiber at any given point, and these two measurement points move along the optical sensing fiber as light travels along the optical fiber. Changes in optical properties may be induced by strain, vibration, acoustic signals and/or temperature as a result of the fluid flow. Phase and intensity based interferometric sensing systems are sensitive to temperature and mechanical, as well as acoustically induced, vibrations. DAS data can be converted from time series data to frequency domain data using Fast Fourier Transforms (FFT) and other transforms, like wavelet transforms, also may be used to generate different representations of the data. Various frequency ranges can be used for different purposes and where low frequency signal changes may be attributed to formation strain changes or fluid movement and other frequency ranges may be indicative of fluid movement. Various techniques may be applied to generate indicators of events related to the generation and/or expansion of shear induced fracture fields during hydraulic fracturing operations. Althoughdepicts the fiber optic cablein the wellbore, a fiber optic cablemay also be positioned in the wellboreto obtain measurements when the wellboreis hydraulically fractured.