Patent Application
20230074047
The present disclosure relates generally to performing wavefield predictions by using wavefront estimations, and more specifically, to performing predictions of Green's functions by using machine learning.
This section is intended to introduce the reader to various aspects of art that may be related to various aspects of the present disclosure, 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 admissions of prior art.
Seismic data can be data that is collected in the course of performing a seismic survey. A seismic survey includes generating an image or map of a subsurface region of the Earth by sending sound energy down into the ground and recording the reflected sound energy that returns from the geological layers within the subsurface region. During a seismic survey, an energy source is placed at various locations on or above the surface region of the Earth, which may include hydrocarbon deposits. Each time the source is activated, the source generates a seismic (e.g., sound wave) signal that travels downward through the Earth, is reflected, and, upon its return, is recorded using one or more receivers disposed on or above the subsurface region of the Earth. The seismic data recorded by the receivers may then be used to create an image or profile of the corresponding subsurface region.
Upon creation of an image or profile of a subsurface region, these images and/or profiles can be used to interpret characteristics of a formation.
A summary of certain embodiments disclosed 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. Indeed, this disclosure may encompass a variety of aspects that may not be set forth below.
A Green's function (G) can generally be considered to be a wavefield solution of an equation LG=δ, where L can be a linear differential operator, and where δ can be a Dirac delta function. The Dirac delta function can be a tool for modelling the physics of a point particle, for example. Green's functions are used as basis functions for building a wavefield. The process of building a wavefield is necessary to perform seismic modeling and inversion.
Certain applications are implemented/obtained from a seismic response, where the seismic response is calculated based on a utilized velocity model, a given source information, and a given receiver information. These applications include applications related to seismic imaging, Full Waveform Inversion (FWI), inversion, illumination, and some post-migration processing, for example. In order for the above-described applications to perform their functions, the applications may need to determine and to utilize the correct/applicable Green's functions. The correct Green's functions can generally be the Green's functions that are applicable to the relevant seismic area of interest.
Further, in order to properly perform their functions, the applications need to repeatedly determine and need to repeatedly utilize the correct Green's functions. The process of determining the correct Green's functions can be computationally costly. In view of the difficulties of determining the correct Green's function, one or more embodiments are directed to a machine learning system that performs the function of learning the correct/applicable Green's functions. The machine learning system can be a deep-learning system, for example.
One or more embodiments of the present invention can generate an estimated wavefront, and one or more embodiments use the estimated wavefront as a guide image, as described in more detail below. One or more embodiments inputs the guide image into the machine learning system, and the machine learning system can predict Green's functions based on the received guide image. In other words, one or more embodiments train the machine learning system to predict/identify Green's functions based on an inputted guide image.
One or more embodiments can generate a guide image based on velocity model information and/or source wave information of a certain seismic area of interest, for example. By generating a guide image (based on velocity model information and/or source wave information), one or more embodiments of the present invention can transform the velocity model information and source wave information into a pattern that can be understood/processed by the machine learning systems. As described above, the estimated waveform serves as a guide image, and use of the guide image can help improve the training of the machine learning (ML) system. With this guide image (i.e., estimated waveform) as an ML input, the neural network underpinning the ML system can quickly provide the output/prediction, which is the applicable Green's function(s) that is determined by the ML system based on the input.
In view of the above, in contrast to other approaches that use inputs (to a ML system) that are expressed in the frequency domain, one or more embodiments of the present invention can use inputs that are expressed in the time domain. As such, in contrast to the other approaches, the present approach does not require wavefield calculations to be performed beforehand.
With one or more embodiments, a method can include receiving at least one wavefield estimation. The method can also include generating an output via at least one machine learning system. The machine learning system can be a deep-learning processor, a classification processor, and/or segmentation processor based on the received wavefield estimation. The method can also include comparing the output of the ML system with a desired output. The method can also include modifying the ML system so that the output corresponds to the desired output, where the desired output can be an applicable/correct Green's function that corresponds to the input.
With one or more embodiments, a method can include receiving at least one wavefield estimation. The received wavefield estimation can be considered to be a guide image. The method can also include generating an output via the at least one trained ML system based on the received wavefield estimation. The output can be a predicted/determined Green's function, for example.
Various aspects of this disclosure may be better understood upon reading the following detailed description and upon reference to the drawings in which:
One or more specific embodiments will be described below. In an effort to provide a concise description of these embodiments, not all features of an actual implementation are 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.
By way of introduction, seismic data may be acquired in the course of implementing a variety of seismic survey systems and techniques, two of which are discussed with respect to
Referring now to
After exploration equipment has been placed within the subsurface region, at block 16, the hydrocarbons that are stored in the hydrocarbon deposits may be produced via natural flowing wells, artificial lift wells, and the like. At block 18, the produced hydrocarbons may be transported to refineries and the like via transport vehicles, pipelines, and the like. At block 20, the produced hydrocarbons may be processed according to various refining procedures to develop different products using the hydrocarbons.
It should be noted that the processes discussed with regard to the method 10 may include other suitable processes that may be based on the locations and properties of hydrocarbon deposits as indicated in the seismic data acquired via one or more seismic survey. As such, it should be understood that the processes described above are not intended to depict an exhaustive list of processes that may be performed after determining the locations and properties of hydrocarbon deposits within the subsurface region.
With the foregoing in mind,
The marine survey system 22 may include a vessel 30, one or more seismic sources 32, a (seismic) streamer 34, one or more (seismic) receivers 36, and/or other equipment that may assist in acquiring seismic images representative of geological formations within a subsurface region 26 of the Earth. The vessel 30 may tow the seismic source(s) 32 (e.g., an air gun array) that may produce energy, such as sound waves (e.g., seismic waveforms), that is directed at a seafloor 28. The vessel 30 may also tow the streamer 34 having a receiver 36 (e.g., hydrophones) that may acquire seismic waveforms that represent the energy output by the seismic source(s) 32 subsequent to being reflected off of various geological formations (e.g., salt domes, faults, folds, etc.) within the subsurface region 26. Additionally, although the description of the marine survey system 22 is described with one seismic source 32 (represented in
In some embodiments, the land-based receivers 44 and 46 may be dispersed across the surface 42 of the Earth to form a grid-like pattern. As such, each land-based receiver 44 or 46 may receive a reflected seismic waveform in response to energy being directed at the subsurface region 26 via the seismic source 40. In some cases, one seismic waveform produced by the seismic source 40 may be reflected off of different geological formations and received by different receivers. For example, as shown in
Regardless of how the seismic data is acquired, a computing system (e.g., for use in conjunction with block 12 of
Referring now to
The processor 64 may be any type of computer processor or microprocessor capable of executing computer-executable code. The processor 64 may also include multiple processors that may perform the operations described below. The memory 66 and the storage 68 may be any suitable articles of manufacture that can serve as media to store processor-executable code, data, or the like. These articles of manufacture may represent computer-readable media (e.g., any suitable form of memory or storage) that may store the processor-executable code used by the processor 64 to perform the presently disclosed techniques. Generally, the processor 64 may execute software applications that include programs that process seismic data acquired via receivers of a seismic survey according to the embodiments described herein.
With one or more embodiments, processor 64 can instantiate or operate in conjunction with a deep-learning processor, a neural-network processor, a classification processor, and/or segmentation processors. With one or more embodiments, the processors can be linear classifiers (such as, for example, Multi-Layer Perception classifiers), support vector classifiers, and/or quadratic classifiers, for example. With another embodiment, the classification and/or segmentation processors can be implemented by using neural networks. The one or more neural networks can be software-implemented or hardware-implemented. One or more of the neural networks can be a convolutional neural network. With one or more embodiments, the classification and/or segmentation processors can perform image segmentation.
With one or more embodiments, these classification and/or segmentation processors can provide responses to different inputs. The process by which a classification and/or segmentation processor learns and responds to different inputs may be generally referred to as a “training” process.
The memory 66 and the storage 68 may also be used to store the data, analysis of the data, the software applications, and the like. The memory 66 and the storage 68 may represent nontransitory computer-readable media (e.g., any suitable form of memory or storage) that may store the processor-executable code used by the processor 64 to perform various techniques described herein. It should be noted that non-transitory merely indicates that the media is tangible and not a signal.
The I/O ports 70 may be interfaces that may couple to other peripheral components such as input devices (e.g., keyboard, mouse), sensors, input/output (I/O) modules, and the like. I/O ports 70 may enable the computing system 60 to communicate with the other devices in the marine survey system 22, the land survey system 38, or the like via the I/O ports 70.
The display 72 may depict visualizations associated with software or executable code being processed by the processor 64. In one embodiment, the display 72 may be a touch display capable of receiving inputs from a user of the computing system 60. The display 72 may also be used to view and analyze results of the analysis of the acquired seismic data to determine the geological formations within the subsurface region 26, the location and property of hydrocarbon deposits within the subsurface region 26, predictions of seismic properties associated with one or more wells in the subsurface region 26, and the like. The display 72 may be any suitable type of display, such as a liquid crystal display (LCD), plasma display, or an organic light emitting diode (OLED) display, for example. In addition to depicting the visualization described herein via the display 72, it should be noted that the computing system 60 may also depict the visualization via other tangible elements, such as paper (e.g., via printing) and the like.
With the foregoing in mind, the present techniques described herein may also be performed using a supercomputer that employs multiple computing systems 60, a cloud-computing system, or the like to distribute processes to be performed across multiple computing systems 60. In this case, each computing system 60 operating as part of a super computer may not include each component listed as part of the computing system 60. For example, each computing system 60 may not include the display 72 since multiple displays 72 may not be useful to for a supercomputer designed to continuously process seismic data.
After performing various types of seismic data processing, the computing system 60 may store the results of the analysis in one or more databases 74. The databases 74 may be communicatively coupled to a network that may transmit and receive data to and from the computing system 60 via the communication component 62. In addition, the databases 74 may store information regarding the subsurface region 26, such as previous seismograms, geological sample data, seismic images, and the like regarding the subsurface region 26.
Although the components described above have been discussed with regard to the computing system 60, it should be noted that similar components may make up the computing system 60. Moreover, the computing system 60 may also be part of the marine survey system 22 or the land survey system 38, and thus may monitor and control certain operations of the seismic sources 32 or 40, the receivers 36, 44, 46, and the like. Further, it should be noted that the listed components are provided as example components and the embodiments described herein are not to be limited to the components described with reference to
In some embodiments, the computing system 60 may generate a two-dimensional representation or a three-dimensional representation of the subsurface region 26 based on the seismic data received via the receivers mentioned above. Additionally, seismic data associated with multiple source/receiver combinations may be combined to create a near continuous profile of the subsurface region 26 that can extend for some distance. In a two-dimensional (2-D) seismic survey, the receiver locations may be placed along a single line, whereas in a three-dimensional (3-D) survey the receiver locations may be distributed across the surface in a grid pattern. As such, a 2-D seismic survey may provide a cross sectional picture (vertical slice) of the Earth layers as they exist directly beneath the recording locations. A 3-D seismic survey, on the other hand, may create a data “cube” or volume that may correspond to a 3-D picture of the subsurface region 26.
In addition, a 4-D (or time-lapse) seismic survey may include seismic data acquired during a 3-D survey at multiple times. Using the different seismic images acquired at different times, the computing system 60 may compare the two images to identify changes in the subsurface region 26.
In any case, a seismic survey may be composed of a very large number of individual seismic recordings or traces. As such, the computing system 60 may be employed to analyze the acquired seismic data to obtain an image representative of the subsurface region 26 and to determine locations and properties of hydrocarbon deposits. To that end, a variety of seismic data processing algorithms may be used to remove noise from the acquired seismic data, migrate the pre-processed seismic data, identify shifts between multiple seismic images, align multiple seismic images, and the like.
After the computing system 60 analyzes the acquired seismic data, the results of the seismic data analysis (e.g., seismogram, seismic images, map of geological formations, etc.) may be used to perform various operations within the hydrocarbon exploration and production industries. For instance, as described above, the acquired seismic data may be used to perform the method 10 of
In some embodiments, the results of the seismic data analysis may be generated in conjunction with a seismic processing scheme that includes seismic data collection, editing of the seismic data, initial processing of the seismic data, signal processing, conditioning, and imaging (which may, for example, include production of imaged sections or volumes (which may, for example, include production of imaged sections or volumes) in prior to any interpretation of the seismic data, any further image enhancement consistent with the exploration objectives desired, generation of attributes from the processed seismic data, reinterpretation of the seismic data as needed, and determination and/or generation of a drilling prospect or other seismic survey applications. As a result, location of hydrocarbons within a subsurface region 26 may be identified. Techniques for detecting subsurface features from the seismic data/images will be described in greater detail below.
If the machine learning system uses a classification and/or a segmentation processor, the classification and/or segmentation processor can be a Multi-Layer Perceptron (MLP) classifier. Although one or more embodiments can use a MLP classifier, other embodiments can use other types of classifiers such as, for example, other linear classifiers, support vector classifiers, quadratic classifiers. The classification and/or segmentation processor can also be implemented using convolutional neural networks (CNNs), and/or recurrent neural networks (RNNs), etc.
As described previously, the computing system 60 having the processor 64 may be any type of computer processor or microprocessor capable of executing computer-executable code and the processor 64 can instantiate or operate in conjunction with a deep-learning processor, a neural-network processor, a classification processor, and/or segmentation processors to perform the operations described in greater detail below.
To use seismic data to produce images, typically seismic migration is utilized to relocate events (e.g., in space or time) to the location that the event occurred in a subsurface region 26 of the Earth rather than at the location that it was recorded at the surface (e.g., the surface 42 of the Earth or marine surface thereof) so as to generate a more accurate image of the subsurface region 26 of the Earth. In seismic migration, for example, reverse time migration (RTM), migration operators (i.e., fundamental solutions to the wave equation) are utilized in the process of generating seismic images to generate a wavefield (e.g., a wavefield from a point source).
A Green's function (G) can generally be considered to be a wavefield solution of an equation LG=δ, where L can be a linear differential operator, and where δ can be a Dirac delta function. In this manner, Green's functions are wavefield solutions for a delta point source. In this manner, Green's functions are used as basis functions for building a wavefield, whereby the process of building a wavefield is allows for performance of seismic modeling and inversion.
Indeed, wavefields can be decomposed by Green's functions. Accordingly, once a Green's function is determined, ware propagation can be predicted. Applications in, for example, seismic imaging, full waveform inversion (FWI), inversion, illumination, and various post-migration processing processes utilize seismic responses from a velocity model given source and receiver information. Each of these instances benefit from Green's functions.
One approach for applying a Green's function is to utilize an approximate expression for the wavefield solution. This can be based on, for example, the travel time of one or more waves. However, this approach can have problems in the accuracy of the result generated. Another approach for applying a Green's function as a waveform solution includes solving a partial differential equation to simulate the wavefield. However, this technique can be computationally costly and difficult to recalculate if one or more input parameters are altered.
A further approach may include of using machine learning, deep learning, and/or neural networks to learn Green's functions as performed in a frequency domain (e.g., attempts to analyze inputs that are in the frequency domain) while still another approach can include learning a time step insights gained by previous time steps. However, these approaches for determining the correct Green's functions can be computationally costly and again can be difficult to modify when desired changes to input parameters are present. Other approaches include the calculation of wavefields for a few time steps beforehand, while another technique involves attempts to learn (e.g., determine) wavefield/Green's functions directly from a velocity model in conjunction with the aforementioned machine learning, deep learning, and/or neural networks to learn Green's functions. This approach is illustrated in
Referring to
The process undertaken by the machine learning system 520 in
Thus, in contrast to the approach outlined above with respect to
The (input) guide image 620 is an approximation of the true the wavefield solution (i.e., an approximation of the Green's function to be generated by a machine learning system). However, through the additional information of the guide image 620 being provided to a machine learning system, this increases the accuracy of the true the wavefield solution generated not only from the velocity model 610, but additional velocity models related to (i.e., velocity models which resemble velocity model 610).
In this manner, the training of the machine learning system is not only applicable to the velocity model 610, but to additional velocity models (i.e., the trained machine learning system can solve for wavefield solutions of differing velocity models). Moreover, once trained, the machine learning system operates more rapidly than a technique of solving for a wavefield solution through, for example, solving for/applying a Green's function as a waveform solution inclusive of solving a partial differential equation to simulate the wavefield. This provides additional benefits of reduced computational (and, accordingly, financial) cost, thus increasing the ease with which the trained machine learning system can be utilized to recalculate a waveform solution if one or more input parameters (e.g., portions of the velocity model 610) are altered. That is, an altered velocity model relative to velocity model 610 can be supplied to the trained machine learning system to generate a waveform solution of the altered velocity model.
The method 800, at step 810, can include receiving a guide image 620 that is to be recognized by a machine learning system 720. The method 800, at step 820, includes generating an output via the machine learning system 720 based on the received guide image 620. The method 800, at step 830, can include comparing the output 730 of the machine learning system 720 with a desired output. This may include checking the output 730 against known results generated independent from the machine learning system 720 (i.e., to check the efficacy of the machine learning system 720). The method 800, at step 840, can also include modifying the machine learning system 720 (or one or more inputs thereto) so that the output 730 corresponds to the desired output, for example, is within a set tolerance with respect to the desired output. This process outlined in method 800 represents training of the machine learning system 720.
Additionally,
The method 900 may represent implementation of a trained machine learning system 720 as trained, for example, through one or more of the steps of method 800 discussed above. The method 900, at step 910, includes receiving a guide image 620. The method also includes, at 920, generating an output 730 via the machine learning system 720 based on the received guide image 620 and using the additional inputs previously discussed with respect to
It should be noted that the steps 910 and 920 can be repeated for additional velocity models using the same guide image 620 so as to create an ensemble of outputs where each unique output is related to its respective input value for a velocity model. This allows for an ensemble of migrations to be undertaken, each having a unique Green's function (output 730) as a migration function. That is, for each velocity model generated for a given migration operation, method 900 can be implemented to generate its Green's function. And when modifications to a generated velocity model are made, method 900 allows for generation of a new corresponding Green's function to be generated therefrom without the need for costly computational analysis for the new velocity model. This allows for generation of an ensemble of seismic images (based on the ensemble of migrations, which themselves are based on an ensemble of velocity models) using the techniques of method 900 much more rapidly and cost efficiently relative to, for example, applying a Green's function as a waveform solution by solving a partial differential equation to simulate the wavefield.
Utilization of the techniques discussed above result in a computing system (e.g., computing system 60) that differs from other computing systems. For example, the training process outlined above for the machine learning system 720 results in a computing system that is different than a computing system having a machine learning system (e.g., machine learning system 520) trained using different inputs. The techniques of utilizing the guide image 620, as described above, in training the machine learning system 720 result in a different computing system having that machine learning system therein because the computing system with the trained machine learning system 720 will generate different resultant outputs than similar systems that have not been trained in the manner described above.
Additionally, having a computer system that includes the trained machine learning system 720 improves the computer capabilities and functionality. As previously noted, the computer system described herein includes a machine learning system 720 that is trained differently than, for example, a machine learning system 520. This training of the machine learning system 720 causes the computer system incorporating the machine learning system 720 to be functionally improved relative to a computer system incorporating the machine learning system 520. Indeed, by providing the machine learning system 720, efficient use of processing power, memory, storage space, network bandwidth, and/or other computing resources is accomplished. This has the dual effect of increasing the efficiency with which users can navigate through seismic imaging processes and thereby making efficient use of processing power, memory, storage space, network bandwidth, and/or other computing resources.
The specific embodiments described above have been shown 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.
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 [perform]ing [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).
This application is a Non-Provisional application claiming priority to U.S. Provisional Patent Application No. 63/241,158, entitled “Method and Apparatus for Performing Wavefield Predictions By Using Wavefront Estimations”, filed Sep. 7, 2021, which is hereby incorporated herein by reference in its entirety for all purposes.
| Number | Date | Country | |
|---|---|---|---|
| 63241158 | Sep 2021 | US |
