EA044564B1 - BUILDING A SPEED MODEL - Google Patents

Description

Link to related applications

This application claims priority to U.S. Provisional Patent Application No. 62/884946, filed August 9, 2019, entitled Construction of a Velocity Model, the entire contents of which are incorporated herein by reference.

State of the art

In general, the present disclosure relates to seismic imaging and, in particular, to seismic imaging used in seismic exploration or seismic monitoring of hydrocarbon production.

This section is intended to familiarize the reader with various aspects of the prior art that may relate to the various features of the present invention described and/or claimed below. It is believed that the present discussion will be useful in familiarizing the reader with information regarding the subject matter of the invention and will provide a better understanding of the various features of the present invention. Accordingly, this information should be perceived from this point of view, and not as a description of the achievements of the state of the art.

Seismic imaging typically involves generating an image or map of the Earth's subsurface by sending sound energy into the geologic environment and recording the reflected sound energy returning from the geologic layers in the subsurface. During seismic surveys, an energy source is placed at various locations on or above the Earth's surface in areas where hydrocarbon deposits may be located. Each time a source is activated, it generates a seismic signal (eg, a sound wave) that travels down underground, is reflected, and is recorded upon return using one or more receivers located on or above the Earth's subsurface. The seismic data recorded by the receivers can then be used to create an image or profile of the corresponding subsurface zone.

Over time, as hydrocarbons are extracted from the Earth's subsurface, the location, saturation, and other characteristics of the hydrocarbon reservoir (eg, cap rock, wellbore interval) within the subsurface may change. Therefore, it is useful to determine how the image or map of the subsurface changes over time in order to introduce changes in hydrocarbon recovery activities to improve the efficiency of hydrocarbon recovery from the Earth's subsurface.

The essence of the invention

The following is a summary of the substance of certain embodiments disclosed herein. It should be understood that the specifications provided serve as a summary of the spirit of these particular embodiments and are not intended to limit the scope of the disclosure. In fact, the present disclosure may cover various features that may not be mentioned below. Seismic imaging is based on a subsurface seismic wave velocity model (eg, velocity model). A velocity model can be a form of representation of geological strata that can be used in seismic data analysis. One method (or family of methods) for modeling seismic wave velocity, called tomography, is based on the analysis of migrated images generated during initial model evaluation and subsequent model updates. Model updating can be based on tomographic back-projection measurements of seismic image kinematic continuity error, typically as a function of depth or time differences in pre-stack migrated images, layered as a function of distance from the shot point or angle. By using acquired data (i.e., time series of recorded data, or seismic traces) having different surface offsets from the shot point (i.e., distances between transmitting sources and receivers) to generate a seismic image, the seismic data can be divided according to the respective distances between the source and the receiver. A speed estimate can be performed and image series generated in which each individual image is created from a subset of data from the entire data set. A subset of data is defined by its constituent seismic tracks, for which the interval of shot point distances between source and receiver is different from that of other subsets. If the velocity estimate is correct, all collected images as a function of their distance along the surface from the point of explosion will correspond to a given (i.e., comparable) depth (i.e., the depth of the reflector under study).

On the other hand, if the collected images are not comparable at a given (ie, comparable) depth (eg, the seismic data at near and far distances from the explosion point are not the same), the estimated velocity model is incorrect in at least some respect. Adjustments to the velocity estimate can be made by using the expected propagation time for the tracks from a common point (i.e., the target reflector) and their surface distances from the explosion point to generate an error estimate of the velocity model. This process can be repeated for different depths or areas, and all velocity model error estimates can be collected and analyzed to determine a new velocity estimate that best fits the generated velocity model error estimates. This process (tomographic) is applicable in cases where the area of the Earth to be imaged is relatively homogeneous (i.e., there are no complicating features such as a volcanic bed or any other complex geological features). In subsurface regions where the true wave speed is subject to large spatial variations (i.e., where the region through which the waves must travel to the target to be imaged is not uniform), it may be difficult to find an accurate sound speed model based on simple methods. Thus, although the first method (tomography) has high computational performance, its accuracy is low, and to update the sound speed model, the method uses simplified wave propagation models, which are effective only if the true model is sufficiently smooth.

Therefore, in relatively complex areas, a more complex method may be required. This second method (or family of methods) for modeling seismic wave velocity, called full waveform inversion (FWI), involves iteratively building a sound speed model using waveform matching, which uses the wave equation and computer modeling to find a sound speed model that produces a data set that approximates the recorded data. This second method performs best with low-frequency waves traveling along a diving path from the source to the receivers.

The advantage of the second method (FWI) over the first method (tomography) is that it is more general and better takes into account the physics of wave propagation. However, since the second method involves attempts to directly model seismic data, a mismatch between real physics and computational physics may lead to incomplete results. Additionally, since the second method does not use seismic migration and image analysis, it does not necessarily provide a better model for seismic migration. Moreover, the second method gives the best results for the shallow subsurface, but may not be effective at greater depths (i.e., diving waves may only be able to reach the regions of interest at depths that are approximately one-third the distance between source and receiver).

Therefore, instead of the waveform matching (FWI) method or the tomographic back projection (tomography) method, a third method (i.e., building a desired velocity model) can be used to model seismic wave velocity. Building a desired speed model directly aims to create a sound speed model that achieves the highest quality sound speed model measured directly from the migrated image. This third method uses the optimality metric of the sound speed model. Using this method, the sum of the squares of the image amplitudes (often called stack power) can be used as a measure of the objective function of the parameter model for seismic processing.

Additionally, this method can estimate stacking power in spatial windows after converting images to normal incidence propagation times via their own sound speed model, to improve the stationarity of the objective function with respect to model changes (e.g., reflectors in an image generated using the updated current speed may move, so when converting images to propagation time at normal incidence, the direction of movement of the reflectors can be changed so that their movement can be compensated to produce a more stationary image for evaluation).

Subsequently, to reduce the dimensionality of the model space and increase the chances of finding the optimal model, building a model of the desired speed includes performing updates to the sound speed model, which is decomposed into a set of spatially localized basis functions of the speed disturbance separated by a multi-level (hierarchical) scale, including a search starting from the initial coarse scale decompositions to finer ones, where updates across fine scales are largely linearly independent from scale to scale. Finally, building a model of the desired velocity involves performing multiple migrations of the data set to find the optimal model, using a robust but efficient prestack migration algorithm (e.g., Kirchhoff, one-way wave equation, reverse time domain migration (RTM). time migration), etc.)

Thus, desiring velocity model construction combines optimized migration with decomposition of the model into a hierarchical set of spatially localized basis functions, and uses stacking power as an effective objective function to find velocity models for seismic migration. Thus, methods for constructing a desired velocity model can be applied to further improve the velocity model derived from tomography and/or FWI.

- 2 044564

Brief description of drawings

Various features of the present disclosure will be better understood by reference to the following detailed description and reference to the drawings, in which: FIGS. 1 is a flow diagram of various processes that can be carried out based on the analysis of seismic data acquired by a seismic acquisition system in accordance with the illustrated embodiments;

in fig. 2 illustrates a first offshore survey system;

in fig. 3 illustrates a first terrestrial imaging system;

in fig. 4 illustrates a computing system suitable for performing the operations described herein based on data acquired by the marine survey system shown in FIG. 2, and/or the ground survey system shown in FIG. 3;

in fig. 5 is a flow diagram of a method for generating a seismic image using the computing system shown in FIG. 4;

in fig. 6 illustrates a second terrestrial imaging system;

in fig. 7 shows a model of the Earth including an image generated using a fixed data set and a trial velocity model;

in fig. Figure 8 shows a graph of velocity disturbances at a point in space, correlated with the distortion area in FIG. 7; and in fig. 9A and 9B show a flow diagram of a method for constructing a model of the desired speed.

Detailed Description of the Invention

The following is a description of one or more particular embodiments. In an effort to provide a brief description of these embodiments, not all practical implementation features are described herein. It should be understood that the development of any such practical implementation, whether in any engineering project or development work, to achieve specific development goals requires the resolution of numerous implementation-specific challenges, such as compliance with various system or commercial constraints, which may vary from one performance to another. Moreover, it should be understood that such development activities may be complex and time-consuming, but will represent a standard set of design, manufacturing, and mass production processes carried out by those skilled in the art with knowledge of this disclosure.

Seismic data can provide valuable information in describing, for example, the location and/or changes of hydrocarbon deposits in the Earth's subsurface. Improvements in seismic data processing and subsequent image generation can be achieved through improvements to the subsurface seismic wave velocity model. In the above discussion, one method (or family of methods) for seismic wave velocity modeling, namely tomography, is based on the analysis of migrated images generated during initial model evaluation and subsequent model updating based on tomographic back projection measurements of seismic image kinematic continuity error, typically depending from differences in depth or time of migrated images to stacking, layered as a function of distance from the point of explosion or corner. The second method (or family of methods) for seismic wave velocity modeling, full wave inversion (FWI), involves iteratively constructing a sound speed model using a waveform matching method that uses the wave equation and computer modeling to find a sound speed model that can create a set data approximating the recorded data. There are, however, cases where both tomography and FWI methods are not sufficiently effective to adequately represent the seismic wave velocity model in the subsurface zone. Accordingly, a third method (or family of methods) may be applied. This third method (building a desired speed model) directly searches for a sound speed model that provides the highest quality score for the sound speed model measured directly from the migrated image. Additionally, methods for constructing a desired velocity model can be applied to further improve the velocity model derived from tomography and/or FWI.

It should be immediately noted that seismic data can be acquired using a large number of different seismic survey systems and techniques, two of which are discussed with reference to FIG. 2 and 3. Regardless of the seismic data acquisition method used, once the data is acquired, the computing system can analyze this data, and can use the results of the seismic data analysis (for example, seismograms, a map of geological productive horizons, etc.) to perform various operations in the field of hydrocarbon exploration and production. For example, in FIG. 1 is a flow diagram of method 10 illustrating various processes that can be carried out based on analysis of acquired seismic data. Although method 10 is described in a specific order, it is noted that implementation of method 10 may be carried out in any suitable order.

- 3 044564

In fig. 1, in step 12, based on the analysis of seismic data, the locations and properties of hydrocarbon deposits in the subsurface of the Earth associated with the corresponding seismic survey can be determined. In one embodiment, seismic data acquired using one or more seismic acquisition techniques may be analyzed to construct a map or profile illustrating various geologic formations in the subsurface.

Based on the identified locations and properties of hydrocarbon deposits, specific locations or portions of the subsurface zone may be explored in step 14. In other words, hydrocarbon exploration organizations can use the locations of hydrocarbon deposits to select drilling sites in the subsurface. In doing so, hydrocarbon exploration organizations can use the locations and properties of hydrocarbon deposits and associated caprock to determine drilling routes, drilling parameters, etc.

After exploration equipment has been installed within the subsurface zone, in step 16, the hydrocarbons stored in the hydrocarbon reservoir can be produced through natural flow wells, artificial lift wells, etc. At step 18, the produced hydrocarbons can be transported to refineries, etc., through vehicles, pipelines, and the like. In step 20, the produced hydrocarbons may be processed through various processing processes to produce various hydrocarbon-using products.

It is noted that the processes discussed with reference to method 10 may include other suitable processes, which may be based on the location and properties of hydrocarbon deposits as determined by seismic data collected through one or more seismic surveys. It can be imagined that the processes described above do not in themselves suggest the existence of an exhaustive list of processes that can be performed once the locations and properties of hydrocarbon deposits in the subsurface have been determined.

With that said, let's look at FIG. 2, a marine survey system 22 (eg, for use in conjunction with pitch 12 of FIG. 1) that can be used to collect seismic data (eg, waveforms) related to the Earth's subsurface in a marine area. In general, marine seismic surveys using the marine survey system 22 may be conducted in the ocean 24 or other body of water above the Earth's subsurface zone 26 underlying the seafloor 28.

The marine survey system 22 may include a vessel 30, a seismic source 32, a marine seismic cable 34, a receiver 36, and/or other equipment that may be useful in collecting seismic images representative of geological horizons in the Earth's subsurface zone 26. Vessel 30 may tow a seismic source 32 (e.g., an array of air guns) that may generate energy, such as acoustic waves (e.g., seismic vibrations) directed at the seabed 28. Vessel 30 may also tow a seismic streamer 34 having a receiver 36 (e.g. , hydrophones) that can receive seismic vibrations (waves) representing the energy of the output vibrations of seismic sources 32 after their reflections from various geological formations (for example, salt domes, fractures, folds, etc.) in the subsurface zone 26. In addition, although described Marine survey system 22 includes one seismic source 32 (represented as an array of air guns in FIG. 2) and one receiver 36 (represented as an array of hydrophones in FIG. 2), it should be noted that marine survey system 22 may include a large number of seismic sources 32 and a large number of seismic receivers 36. Likewise, although the marine survey system 22 described above includes a single marine seismic cable 34, it is noted that the marine survey system 22 may include a large number of seismic cables 34. In addition, additional vessels 30 may have additional seismic cables. sources 32, seismic streamers 34, etc., for the operation of the survey system 22.

In fig. 3 illustrates a terrestrial survey system 38 (eg, for use in conjunction with pitch 12 of FIG. 1) that can be used to obtain information related to the Earth's subsurface zone 26 for non-marine environments. The surface acquisition system 38 may include a (terrestrial) seismic source 40 and a (terrestrial) receivers 44. In some embodiments, the surface acquisition system 38 may include one or more multibeam seismic sources 40 and one or more seismic receivers 44 and 46. For purposes of discussion at fig. 3 shows a seismic source 40 and two seismic receivers 44 and 46. The seismic source 40 (e.g., a seismic vibrator) may be located on the Earth's surface 42 above the subsurface region of interest 26. The seismic source 40 may generate energy (e.g., acoustic waves, seismic vibrations) , aimed at the subsurface zone 26 of the Earth. Upon reaching various geological formations (e.g., salt domes, fractures, folds, etc.) in the subsurface zone 26, the output energy of the seismic source 40 may be reflected from these geological formations and collected or recorded by one or more surface receivers

- 4 044564 (for example, 44 and 46).

In some embodiments, seismic receivers 44 and 46 may be distributed over the Earth's surface 42 to form a grid structure. Each seismic receiver 44 or 46 may receive a reflected seismic waveform in response to energy directed into the subsurface zone 26 from the seismic source 40. In some cases, a single seismic waveform generated by the seismic source 40 may be reflected from different geological formations and may be received different receivers. For example, as shown in FIG. 3, the seismic source 40 may generate an output wave whose energy is directed to the subsurface zone 26 in the form of a seismic wave 48. The first receiver 44 may receive the reflection of the seismic wave 48 from one geological formation, and the second receiver 46 may receive the reflection of the seismic wave 48 from another geological formation. education. In other words, the first seismic receiver 44 may receive the reflected seismic waveform 50, and the second receiver 46 may receive the reflected seismic waveform 52.

Regardless of how the seismic data was collected, the computing system (for example, for use in conjunction with step 12 in FIG. 1) can perform an analysis of the seismic waves received by seismic receivers 36 (offshore) or seismic receivers 44 and 46 ( ground-based), to determine information related to the geological structure, location and properties of hydrocarbon deposits, etc. within the subsurface zone 26. In FIG. 4 illustrates an example of such a computing system 60 that can perform various data analysis operations to examine seismic data collected by receivers 36, 44, 46 to determine the structure of geological formations in the subsurface zone 26.

As shown further in FIG. 4, the computing system 60 may include a communications component 62, a processor 64, a storage device 66 (e.g., a tangible, non-volatile, computer-readable media), a data storage device 68 (e.g., a tangible, non-volatile, computer-readable medium), input/output ports 70 (I /O), display 72, etc. Communications component 62 may be a wireless or wired component that may provide communication between receivers 36, 44, 46, one or more databases 74, other computing devices, and/or other devices having communication capability. . In one embodiment, the computing system 60 may receive data 76 (e.g., seismic data, seismograms) that may previously have been received by seismic receivers through a network component, database 74, or the like. A processor 64 of the computing system 60 may analyze or process the received data 76 to determining various features related to geological formations in the subsurface zone 26 of the Earth.

Processor 64 may be a computer processor or microprocessor of any type capable of executing a computer program or instructions to implement the methods described herein. Processor 64 may also include multiple processors that may perform the operations described below. The memory device 66 and data storage device 68 may be any suitable off-the-shelf product that can serve as a storage medium for processor-executable computer-readable code, data, etc. These finished products may be a computer-readable medium (eg, any suitable form for storing and storing data) that can store a processor-executable program used by the processor 64 to implement the methods disclosed herein. In general, processor 64 may execute software applications that include programs for processing seismic data collected by seismic receivers in accordance with the embodiments described herein.

Memory device 66 and data storage device 68 may be used for data storage, data analysis, software applications, etc. The memory device 66 and data storage device 68 may be a tangible, non-volatile medium capable of being directly written to and read from data, or a non-volatile computer-readable medium (e.g., a memory or data storage device in any suitable form) that can store a program executable by the processor, used by processor 64 to implement various methods described in the present disclosure. It can be noted that materiality and energy independence simply means that the medium is a material carrier and not a signal.

The I/O ports 70 may be interfaces that may connect to other peripheral components, such as input devices (eg, keyboard, mouse), sensors, input/output (I/O) modules, and the like. The I/O ports 70 may allow the computing system 60 to communicate with other devices in the marine survey system 22, land survey system 38, or others through the I/O ports 70.

The display 72 may display images associated with software or executable programs being processed by the processor 64. In one embodiment, the display 72 may be a touchscreen display that allows input to the user of the computing system 60. The display 72 may also be used to view and examine analysis results. collected seis

- 5 044564 mic data to determine the geological formations within the subsurface zone 26, the location and properties of hydrocarbon deposits in the subsurface zone 26, etc. The display 72 may be any suitable type of display, such as a liquid crystal display (LCD), plasma display, or organic light-emitting diode (OLED) display. It may also be noted that in addition to the display 72 mentioned herein, the computing system 60 may also display through other media, such as paper (printing), etc.

With that said, the method described herein can also be implemented using a supercomputer using multiple computing systems 60, a cloud computing system, or other means to distribute processes across a large number of computing systems. In this case, each computing system 60 operating as part of a supercomputer may not include every component referred to as part of the computing system 60. For example, each computing system 60 may not include a display 72 because the display 72 may not be useful for the supercomputer intended for continuous processing of seismic data.

After performing various processing on the seismic data, the computing system 60 may store the analysis results in one or more databases 74 . Databases 74 may be coupled by signaling to a network that may communicate with computing system 60 via communications component 62. Additionally, databases 74 may store information related to subsurface zone 26, such as previously acquired seismograms, geological sample data, seismic images, etc., related to the subsurface zone 26.

Although the components described above have been discussed in relation to the computing system 60, it should be noted that the computing system 60 may be composed of similar components. Moreover, the computing system 60 may be part of a marine survey system 22 or a land survey system 38, and may monitor and control certain operations of seismic sources 32 or 40, receivers 36, 44, 46, etc. Additionally, it may be noted that the listed the components are given by way of example, and the embodiments described herein are not limited to the components described with reference to FIGS. 4.

In some embodiments, the computing system 60 (specifically, the processor 64 operating in conjunction with at least a memory device 66 or a data storage device 68) may generate a two-dimensional representation or a three-dimensional representation of the subsurface zone 26 based on seismic data acquired through the above-mentioned receivers. . In addition, seismic data associated with multiple source/receiver combinations may be combined to create a nearly continuous profile of the subsurface region 26 that may extend over some distance. In a two-dimensional (2-D) seismic survey, receiver locations may be located along a single line, while in a three-dimensional (3-D) survey, receiver locations may be distributed across the surface in a grid pattern. In this case, 2-D seismic imaging can provide a cross-sectional picture (vertical slice) of the layers of the Earth located directly below the recording points. A 3-D seismic survey, in contrast, can produce a cube or volume of data that can correspond to a 3-D picture of the subsurface zone 26.

In addition, a 4-D seismic survey (periodic survey) may include seismic data acquired by multiple 3-D surveys. Using different seismic images acquired at different points in time, the computing system 60 can compare the two images to detect changes in the subsurface zone 26.

In any case, a seismic survey can be composed of a very large number of individual seismic records or seismic traces. In this case, the computing system 60 may be used to analyze the collected seismic data to obtain an image of the subsurface zone 26 and to determine the location and properties of hydrocarbon deposits. To achieve this, various seismic data processing algorithms can be used to remove noise from acquired seismic data, transfer pre-processed seismic data, identify offsets between a large number of seismic images, register a large number of seismic images, etc.

After the computer system 60 has analyzed the collected seismic data, the results of the seismic data analysis (eg, seismograms, seismic images, geological formation maps, etc.) can be used to perform various operations in hydrocarbon exploration and production. In some embodiments, the computing system 60 may provide an indication of the presence of hydrocarbons. Accordingly, the computing system 60 may indicate a subsurface zone 26 that may potentially contain hydrocarbons and provide the location (e.g., coordinates or relative locations) of areas containing hydrocarbon deposits and/or (in some cases) areas with possible complications in the subsurface drilling process. . In other embodiments, the image generated using the present method may be displayed on a display 72 of the computing system 60.

- 6 044564 facilitating location determination by the user of the computing system 60. Accordingly, the collected seismic data can be used to implement the method 78 of FIG. 5, which details various processes that can be performed based on the analysis of the collected seismic data to perform seismic data analysis.

In some embodiments, a seismic image may be generated in a seismic data processing circuit, such as the method 78 illustrated in FIG. 5, by a computing system (specifically, a processor 64 operating in conjunction with at least a memory device 66 or a data storage device 68). As shown in the flow diagram, method 78 includes a series of operations comprising seismic data acquisition at step 80, seismic data editing at step 82, initial processing at step 84, and processing, normalization and display (which may, for example, include creating section displays and volumes) of the signal in step 86 prior to any interpretation of the seismic data, any further enhancement of the image in accordance with specified exploration objectives, generation of attributes from the processed seismic data, reinterpretation of the seismic data if necessary, and determination and/or development of drilling plans and/or or other seismic applications. As a result of the method 78, the locations of hydrocarbons in the subsurface zone 26 can be determined. Detecting the location of hydrocarbons contained in hydrocarbon reservoirs can be based, at least in part, on the execution of the processes in step 86, including pre-stack subsurface migration techniques, using, for example, terrestrial common image gather (CIG) gathers, where prestack depth migration methods use the results of the correction of the applied velocity model used in calculating the prestack depth migration data.

In fig. 6 is an illustration of a terrestrial survey system 88 (although the example shown may also be applicable to a marine system, e.g., instead of a marine survey system 22) that can be used to obtain information related to the Earth's subsurface zone 26, where the subsurface zone 26 includes one or more complex geological features 90 and 92, in addition to relatively homogeneous reflectors 94. As shown in the terrestrial geological system 88, the inclusion of complex geological features 90 and 92 may limit the use of tomography, since waves passing through complex geological features 90 and 92 vibrations 96 are subject to distortion.

In addition, the depth in the Earth of complex geological features 90 and 92 may be too great to allow FWI to be used properly because diving waves 98 penetrating through complex geological features 90 and 92 are not received by receivers 44 (i.e., diving waves 98 can only reach the regions of interest at a depth equal to approximately one-third of the distance between source 40 and receiver 44). Thus, methods for constructing a desired velocity model may be used in lieu of waveform matching (FWI) or back projection tomography methods to model seismic wave velocity, for example in a surface system 88 comprising one or more complex geological features. 90 and 92 (i.e., where the existence and/or location of one or more complex geological features 90 and 92 may be generally known, but their exact location and composition are not known and/or determined by FWI or tomography).

Building a desired velocity model typically involves generating an image as a function of the velocity model (where the desired velocity model is initially unknown, so an initial guess is used as the primary velocity model) and data whose value is fixed (i.e., the recorded data as a fixed set of data). The squares of all image values can be summed to generate an image quality score. This technique may be referred to as summation power optimization or more generally as summation power or summation power measurement. In other words, the quality of an image generated using a fixed data set and a trial speed model (eg, the original speed model) can be predicted based on the sum of the squared values of the generated image. This image quality is an indicator of the optimality of the speed model and will be discussed below as a method used in conjunction with building the desired speed model.

In fig. 7 illustrates an Earth model 100 including an image 102 generated using a fixed data set and a trial (eg, raw) velocity model, as discussed generally with reference to FIG. 6. Image 102 includes a distortion object 104 (eg, a velocity distortion object) corresponding to a complex geological region (eg, a complex geological feature 90 or other distortion object/formation in the subsurface). In addition, reflectors 105 are shown, as well as distorted portions 107 of the image 102 that are generated when the velocity distribution in the distorting object is incorrect.

The Earth layer containing the distorting object 104 is decomposed into a series of basis velocity functions. A velocity disturbance (Δν) can be input into the velocity model and the result used in seismic migration to create an image. The result will be an image of the data using the velocity model perturbed at the jth point in space (i.e., the distance according to the velocity model) with the selected amplitude of the velocity disturbance. The size (i.e., large or small) and sign (i.e., positive or negative) of the velocity disturbance, as well as the point of its position (i.e., the jth point in space, either along the x-axis, xy coordinate axis, or xyz coordinate axis), for example, by random or quasi-random methods, or predefined position points and disturbances can be used.

In fig. Figure 8 shows velocity disturbances 106, 108, 110, 112, 114 and 116 of various sizes and signs. These randomly selected velocity perturbations 106, 108, 110, 112, 114 and 116 are located at the jth point 118 in the space (ie, measured along the x-axis) of the velocity model. The image generation process may be performed independently for data including each of the velocity disturbance data 106, 108, 110, 112, 114, and 116, i.e., an image is generated based on the velocity disturbance data 106, an image is generated based on the velocity disturbance data 108, and etc. The squares of all values of each generated image are summed, i.e., its summation power (S) is determined to generate a quality indicator for this image.

This process can be repeated at additional points in space (i.e., along the x-axis in the velocity model) so that additional velocity perturbations can be mapped at point jth+1 and a corresponding S generated for the velocity perturbations at point jth+1. As noted earlier, the choice of spatial points and/or additional velocity perturbations can be determined in advance or, for example, by random or quasi-random sampling of model samples. This method may, for example, be similar to the Monte Carlo method, or this method may be used.

In this way, the distorting object 104 is broken down into small elements (points), and a sequence of velocity disturbances is applied at selected points in space in the velocity model. At each point, images are generated for each velocity perturbation at that point, and the values of the image squares are summed for each image. This S value is stored in memory. A table or other array of input data S is generated. Thus, for each point at which the velocity model is perturbed (independent of any other point), it is determined and stored, with S being a function of the magnitude of the disturbance and the position of the disturbance point X (for example, at central point Xc of the disturbance). The disturbance can be described by the following basis function of the velocity disturbance

where Xc is the central point of the disturbance (as shown in Fig. 8), Xw is the width scale of the disturbance (shown in Fig. 8) and X is the spatial coordinate in the velocity model (as shown in Fig. 8). The optimal (e.g., highest) values of S are selected (e.g., the optimal speed disturbance at some point, having, for example, the highest value of S) so that a set of optimal values of S is generated. Thus, for each point on the distorting object 104 there is a best estimate velocity model (i.e., changing the velocity model with a particular velocity perturbation) for that point, and these best estimates are treated as an overall update to the velocity model. This velocity model update is generated from perturbations of parts of the velocity model, which are then combined together using the above methods (i.e., individual disturbances at different points in the velocity model are independent of each other and can be combined to determine the complete velocity model update). Therefore, the velocity model update is generated by summing the optimal disturbances (determined by the value of their corresponding S value).

In fig. 9A and 9B illustrate a method 120 for constructing a desired velocity model for use with the ground system of FIG. 6 (or offshore systems with similar complex seismic parameters). Method 120 may be implemented by a computer system 60 (and, in particular, a processor 64 operating in conjunction with at least one of memory devices 66 or storage device 68, e.g., by executing a program or instructions to perform the steps of method 120) . At step 122, a primary speed model is selected and the current speed model is taken to be the primary speed model. As shown above, the desired speed model may not be initially known, so some embodiments use an initial guess as the primary speed model. Next, at step 122, the initial disturbance points and scales are selected. As shown in the block diagram, the selected perturbation starting points include selecting the central perturbation point along the x-axis (Xc), y-axis (Yc), and z-axis (Zc) when the 3D model is generated. Alternatively, one or more of these points may be omitted when the 1-D or 2-D velocity model is generated by method 120. Additionally, the example illustrated at step 122 relates to the choice of scales, eg, x-axis (Xw) and y-axis (Yw) disturbance widths.

At step 124, the current velocity model (in this example, the primary velocity model from step 122 as the current velocity model) is decomposed into a series of basis velocity functions for each study.

- 8 044564 blowing direction (coordinate axis) using the corresponding selected scales. Additionally, at step 124, the maximum and minimum speed disturbance corrections (dvmin and dvmax) are determined. At step 126, the current velocity model is subjected to a selected velocity disturbance at a selected point. At step 128, a migration is performed in conjunction with a display of data including the selected velocity disturbance at the selected point. At step 130, the squares of all values of the image generated in step 128 are summed to generate a quality score, ie, an S value for that image, which is stored or otherwise recorded. At step 132, a determination is made whether there is a need to calculate additional velocity disturbance at the selected point. If additional processing of the velocity disturbance at the selected point is required, a new velocity disturbance is selected in step 134, and steps 126, 128, 130, and 132 are repeated for the other velocity disturbance selected in step 134. If no additional processing of the velocity disturbance at the selected point is required, step 132 also includes determining optimal (ie, highest) values of S (eg, the optimal velocity disturbance at a selected point having the highest S value) based on previous iterations using different velocity disturbances. This found optimal value of S is stored or otherwise stored in memory as the best estimate of the speed model (i.e., the change in the speed model under a particular speed disturbance) for the selected point, for later use in method 120 as part of a full model update speed.

Once the optimal S value for a particular (selected) point is stored or otherwise recorded in step 132, step 136 determines whether the selected point used in conjunction with step 132 is the last point in space at which S values for the velocity model are to be determined. In other words, step 136 determines whether there is an additional point for which the desired speed model needs to be built. If so (i.e., if one or more velocity disturbances at another point of interest are to be processed), the next point is selected in step 138, and steps 126, 128, 130, 132, and 134 are repeated (if necessary), returning to step 136, which again determines whether the previously selected point used in step 132 was the last point in the space for determining S values for the velocity model. If it is determined in step 136 that the previously selected point from step 132 was the last point in the space for determining S values for the velocity model, method 120 then proceeds to step 140.

At step 140, the current velocity model is updated with the sum of the received disturbances (ie, the optimal values of S stored or otherwise recorded at step 132 for each point). In this case, the aggregate of the best estimates of the speed model (i.e., changes in the speed model with a specific speed disturbance at specific points) is taken as a whole as an update of the speed model, i.e., the current speed update is generated by summing the optimal disturbances (defined by their respective values S ) at step 140. At step 142, a determination is made whether the quality of the generated image (eg, in one of steps 140 and 142) using the updated current speed model from step 140, exceeds the threshold value (or, for example, if the full number of iterations, or, for example, if the change in the updated current velocity model from step 140 is less than a threshold value different from the previously generated updated current velocity model from step 140). If the quality of the generated image is determined to be satisfactory at step 142, the process stops at step 144. Otherwise, if the quality of the image generated at step 142 is determined to be unsatisfactory, the process continues at step 146.

At step 146, it is determined whether the maximum or minimum changes in velocity disturbances (dvmin and dvmax) should be narrowed or expanded to improve the quality of the image to be generated at step 140. If it is determined that the maximum and minimum changes in velocity disturbances (dvmin and dvmax ) must be narrowed or expanded in step 146, these new dvmin and dvmax values are used in step 124, and the process described above involving steps 124, 126, 128, 130, 132, 134 (if necessary), 136, 138 (if necessary ), 140 and 142 are repeated. If it is determined that the maximum and minimum changes in velocity disturbances (dvmin and dvmax) should not contract or expand at step 146, the process continues at step 148 in which it is determined whether the scale should be reduced, for example, the width scales of the disturbances along the x-axis ( Xw) and along the y-axis (Yw). If it is determined that the scales should be reduced in step 148, then the new scales are used in step 124 (e.g., Xw and Yw) and the process described above is repeated, including steps 124, 126, 128, 130, 132, 134 (if necessary) , 136, 138 (if necessary), 140 and 142. If it is determined that the scales should not be scaled down in step 148, the process ends in step 150. It is noted that step 148 covers the process of method 120 in which the selected coarse scale basis function is specified, and, using a mathematical annihilation or incremental optimization modeling method, searches for an improved model; when a model is found that cannot be improved at a given scale, the next smaller scale is selected and the iteration is performed again. This process stops at step 150 when a satisfactory model is found, as determined at step 148.

- 9 044564

It is noted that the performance of method 120 can be further improved by estimating the stacking power in a spatial window (i.e., the region of interest, e.g., distorting object 140) after converting the images to normal incidence propagation time through their own sound speed model, to improve stationarity objective function relative to changes in the velocity model. In other words, the position of the reflectors, such as the reflectors in the image generated in step 140 using the updated current speed, may change. Therefore, when converting images to propagation time at normal incidence, the direction of movement of the reflectors is changed so that their movement can be compensated for (ie, a more stationary image is obtained for evaluation in step 142).

It is further noted that if a model is simply decomposed into 3D image elements - voxels (for example, 30 pixels), the space of possible models is astronomically large, and by directly searching for all possible models there is little chance of identifying the optimal model. Accordingly, to reduce the dimensionality of model space and increase the chances of identifying an optimal model, updates to the sound speed model are decomposed into a set of multiscale (hierarchical) scaled spatially localized basis functions of the speed perturbation. For example, to generate the basis functions of the velocity disturbances at step 124, Hanning cutting functions can be used, in which the centers of the basis functions of the velocity disturbances are separated by half a period and the scale is halved when moving to the next scale. The relative independence of the influence of the basis functions of adjacent velocity disturbances in space on the migrated images, and the ability to search for patterns ranging from coarse scales to fine scales, can be used as parameters in generating the basis functions of the velocity disturbances at step 124 (i.e., after perform step 148).

Indeed, the basis function decomposition of the velocity perturbation described here may affect the migrated images. These influences are sufficiently independent from one spatial window to the next that a search method based on mathematical anniling modeling techniques with fast convergence quickly reveals improved sound speed models at a given decomposition scale. Accordingly, building a model of the desired rate can benefit from a search starting with a coarse-scale decomposition and moving to increasingly finer-scale decompositions where updates become essentially linearly independent as one moves between scales. Thus, by initially selecting large velocity perturbations in method 120 and then moving to increasingly smaller velocity perturbations for given points in the second and (if necessary) subsequent iterations in method 120, multilevel functional decomposition (i.e., simulation) can be implemented method of mathematical anniling) to reduce the total volume of operations performed to search for models.

Additionally, this multilevel functional decomposition may be combined using additional search algorithms, such as a Monte Carlo search, to select points for introducing velocity disturbances in step 126. It is further noted that velocity disturbances at a given point generally do not affect other velocity disturbances at additional points.

Additionally, since method 120 may involve multiple database migrations to find the optimal rate model, a high quality but efficient pre-stack migration algorithm is used. Examples of migration algorithms that can provide fast, accurate results include Kirchhoff, one-way wave equation, reverse time migration (RTM), etc. These migration algorithms can be initially applied to low-frequency versions of the database to discover the approximate optimal model and, once found, the frequency content of the data can be expanded to the full frequency spectrum, which can provide fast calculations in the early stages of model refinement, similar to full-wave inversion (FWI). ).

The specific embodiments described above have been presented by way of example, and it should be understood that these embodiments may be subject to various modifications and implementation in other forms. It should also be understood that the claims are not intended to be limited to these specific forms disclosed, but rather are intended to cover all modifications, equivalents and alternatives that fall within the spirit and scope of the present disclosure.

The methods presented and claimed herein relate to and are applicable to tangible objects and specific examples of a practical nature that obviously advance the technical field and, for this reason, are not abstract, intangible or purely theoretical. In addition, if the appended claims at the end of this specification contain one or more elements designated as means for [performing] [function] or a step for [performing] [function], it is intended that such elements are to be construed within the meaning of 35 U.S.C. steam. 112(f). However, for any claims containing elements designated in any other way, it is intended that such elements shall not be construed under 35 U.S.C. par. 112(f).

-

Claims (18)

CLAIM 1. A computer-implemented method for generating a seismic image, in which seismic data is obtained including data corresponding to reflections from a target located under a velocity-distorting object in a subsurface zone, wherein the presence of the velocity-distorting object results in distortion of the seismic image; inputting a velocity disturbance into a current velocity model representing a portion of the subsurface zone responsible for distorting the seismic image to generate a velocity disturbance model; generate an image by migrating seismic data and a perturbed velocity model; generating and assigning a quality score to the image, wherein the quality score assigned to the image is based on the sum of squares of several image values; determining whether the quality score assigned to the image is an optimal quality score at a particular point in the current velocity model corresponding to the portion of the subsurface zone responsible for distortion of the seismic image; and updating the current velocity model used to generate the seismic image to generate a refined velocity model using a quality score assigned to the image, when the quality score assigned to the image is determined to be an optimal quality score at a specific point of the current velocity model corresponding to the portion of the subsurface zone responsible for distortion of seismic images. 2. The method of claim 1, comprising determining whether to introduce an additional velocity disturbance into the current velocity model that represents the portion of the subsurface responsible for distorting the seismic image to generate a second velocity disturbance model. 3. The method of claim 2, wherein a second velocity disturbance is introduced into the current velocity model that represents the portion of the subsurface zone responsible for seismic image distortion to generate the second velocity disturbance model when it is determined that additional velocity disturbances must be introduced into the current velocity model. speed disturbances. 4. The method according to claim 3, wherein a second image is generated using a second perturbed velocity model; and generating and assigning a second quality score to the second image. 5. The method of claim 4, comprising determining whether the second quality score assigned to the second image is an optimal quality score at a particular point in the current velocity model corresponding to a portion of the subsurface zone responsible for distortion of the seismic image. 6. The method of claim 5, wherein the current velocity model used to generate the seismic image is updated to generate a refined velocity model using a second quality score assigned to the second image when the second quality score assigned to the second image is determined to be an optimal quality score at a specific point in the current velocity model, which corresponds to the part of the subsurface zone responsible for the distortion of the seismic image. 7. The method of claim 1, comprising determining whether an additional velocity disturbance should be introduced into the current velocity model at a second point of the current velocity model. 8. The method of claim 7, wherein a second velocity disturbance is introduced into the current velocity model at a second point to generate a second velocity disturbance model when it is determined that an additional velocity disturbance at a second point of the current velocity model is to be introduced into the current velocity model. 9. The method of claim 8, comprising generating a second image using a second perturbed velocity model; and generating and assigning a second quality score to the second image. 10. The method of claim 9, comprising determining whether the second quality score assigned to the second image is a second optimal quality score at a second point of the current velocity model corresponding to a portion of the subsurface zone responsible for distortion of the seismic image. 11. The method of claim 10, wherein the current velocity model used to generate the seismic image is updated to generate a second refined velocity model using a second quality score assigned to the second image when the second quality score assigned to the second image is determined to be the second optimal a quality indicator at the second point of the current velocity model, which corresponds to the portion of the subsurface responsible for seismic image distortion. 12. The method of claim 11, wherein an updated current velocity model is generated by summing the refined velocity model and the second refined velocity model. - 11 044564 13. The method of claim 12, comprising generating a second seismic image using the updated current velocity model, wherein the second seismic image represents hydrocarbons in the Earth's subsurface or subsurface drilling complications; and determining whether the quality of the second seismic image exceeds a threshold value. 14. The method of claim 13, comprising downscaling the width of the disturbance along at least one coordinate axis after determining that the quality of the second seismic image does not exceed a threshold value; updating the current speed model with the updated current speed model; introducing a third velocity disturbance into the current velocity model to generate a third disturbance velocity model; generating a third image using a third perturbed velocity model; generating and assigning a third quality score to the third image; determining whether the third quality score assigned to the third image is the third optimal quality score at a particular point in the current speed model; and updating the current velocity model used to generate the seismic image to generate a third refined velocity model using a third quality score assigned to the third image when it is determined that the third quality score assigned to the third image is the third optimal quality score at a particular point in the current model. speed. 15. A non-volatile computer-readable storage medium containing commands written on it, the execution of which by the processor ensures the acquisition of seismic data, including data corresponding to reflections from a target located under a velocity-distorting object in the subsurface zone, and the presence of a velocity-distorting object leads to distortion of the seismic image ; introducing a velocity disturbance into the current velocity model representing the portion of the subsurface responsible for seismic image distortion to generate a velocity disturbance model; image generation via seismic migration of seismic data and perturbed velocity model; generating and assigning a quality score to the image, wherein the quality score assigned to the image is based on the sum of the squares of multiple image values; determining whether the quality score assigned to the image is an optimal quality score at a particular point in the current velocity model corresponding to the portion of the subsurface zone responsible for distortion of the seismic image; and updating the current velocity model used to generate the seismic image to generate a refined velocity model using a quality score assigned to the image, when the quality score assigned to the image is determined to be an optimal quality score at a specific point of the current velocity model corresponding to a portion of the subsurface zone responsible for distorting the seismic image. 16. The computer-readable storage medium of claim 15, comprising instructions that, when executed by a processor, cause a second velocity disturbance to be introduced into a current velocity model representing the portion of the subsurface responsible for distorting the seismic image to generate a second velocity disturbance model when it is determined that current speed model, an additional speed disturbance must be introduced; generating a second image using the second perturbed velocity model; generating and assigning a second quality score to the second image; and determining whether the second quality score assigned to the second image is an optimal quality score at a particular point in the current velocity model corresponding to a portion of the subsurface zone responsible for distortion of the seismic image. 17. The computer-readable storage medium of claim 16, comprising instructions that, when executed by a processor, cause the current velocity model used to generate the seismic image to be updated to generate a refined velocity model using a second quality score assigned to the second image, when the second quality score assigned to second image, is defined as the optimal quality indicator at a specific point in the current velocity model corresponding to the part of the subsurface zone responsible for the distortion of the seismic image. 18. The computer-readable storage medium of claim 15, comprising instructions that, when executed, cause the processor to determine whether an additional velocity disturbance should be introduced into the current velocity model at a second point of the current velocity model; -