ESTIMATING THE SPEED OF TRAINING BY VERTICAL SEISMIC PROFILING

专利摘要:
The invention relates to a method for processing vertical seismic profiling data (PSV). The method includes receiving PSV data in response to seismic energy applied to the formation, processing a descendant portion of the PSV data associated with a downgoing wave field, generating a first set of estimate values according to the processing of the falling part of the PSV data, the first set of estimation values estimating at least one of the slowness or velocity, processing of an ascending part of the PSV data associated with an ascending wavefield generating a second set of estimate values according to the processing of the upstream portion of the PSV data, the second set of estimate values estimating at least one of the slowness or velocity, and the determination of an estimate associated with the training based on the first and second sets of estimate values.
公开号:FR3069930A1
申请号:FR1856079
申请日:2018-07-02
公开日:2019-02-08
发明作者:Amit Padhi；Mark Elliott Willis
申请人:Halliburton Energy Services Inc；
IPC主号:

专利说明:

OFFICE OF PATENTS AND TRADEMARKS OF THE UNITED STATES
Patent application concerning:
ESTIMATING THE SPEED OF A VERTICAL SEISMIC PROFILING TRAINING
Inventors: Mark Elliott Willis; Amit Padhi
TECHNICAL FIELD OF THE INVENTION The embodiments described here generally relate to the use of vertical seismic profiling (PSV) to obtain an estimate of the speed of a formation and, more particularly, to methods of processing PSV data. say zero offset (zero vertical offset seismicprofilinglVSVZO) using several data sets to estimate the speed of a training.
BACKGROUND OF THE INVENTION Hydrocarbons, such as oil and gas, are generally obtained from underground formations which can be either on land (onshore) or at sea (offshore). The implementation of underground operations and the processes involved in extracting the hydrocarbons present in an underground formation are complex. In general, underground operations involve a number of different steps such as drilling a well through and / or in the underground formation at a desired site, treating the wellbore to optimize the production of 'hydrocarbons and carrying out the steps necessary to produce and process the hydrocarbons extracted from the underground formation. All or part of these steps may require and use seismic / acoustic measurements and other detection data to determine the characteristics of the formation, hydrocarbons, equipment used in operations, etc.
An example of a technique for obtaining seismic / acoustic data includes the use of PSV. The PSV consists in measuring in a wellbore the seismic / acoustic energy coming from a seismic source located on the surface of the formation (truck generating shock waves, air gun, drop in weight and / or explosives, eg.). Traditionally, measurements based on PSV (and generating PSV data) involve sampling a seismic wave field using a chain of seismic / acoustic receivers (e.g. geophones and / or hydrophones). ) approximately evenly spaced, which is lowered into a well. The sampling of a seismic wave field carried out within the framework of the PSV using geophones or hydrophones does not generally make it possible to obtain a resolution greater than several tens of feet.
Another PSV data collection method may include the use of distributed acoustic sensing / OAS techniques. In the case of the PSV using the DAS, not geophones or hydrophones are deployed in the wellbore, but a fiber optic cable. Compared to the PSV based on the use of geophones or hydrophones, the PSV using the DAS allows a simplified deployment which does not interfere with the operations carried out in the wellbore; it also allows the acquisition of instantaneous measurement data over the entire length of the well and improves the resolution. The ability to improve the orientation of the data obtained by seismic profiling, in particular for the DAS-based PSV, is also of direct interest for the extraction of the hydrocarbons contained in the underground formations.
Zero offset vertical seismic profiling (PSVZO) designates a PSV technique in which data collection takes place while the seismic source is arranged near the wellbore, for example, directly above that -this. PSVZO is possible in areas where the geological structure is flat and in strata. The speed of a formation is usually estimated using a single set of data associated with the falling wave field. We use current algorithms to choose a first pause time for each receiver (corresponding to the time it takes for a wave to propagate directly from the source to the receiver) and we determine a slope of the first pause, the slope indicating the associated time shifts slow training (the inverse of the speed of training). If a seismic source containing mainly compression waves (P waves) is used, then the speed of the P waves of the formation can be estimated from the choices of first pause.
Alternatively, if a seismic source containing mainly shear waves (S waves) is used, then the speed of the shear waves of the formation can be estimated from the choices of first break. Thus, an estimate of the speed of the formation can be calculated from the slope determined for the first breaks of the falling wave field. However, the rising reflected wave field, which is subject to the same time lags indicating the speed of the formation, but is also more subject to noise than the falling wave field, is generally not used to estimate the speed d 'a training. The PSV data associated with the rising wave field remains unused to estimate the speed of a formation. Other data not used to estimate the speed of a training include PSV data associated with the rising or falling wave fields associated with time windows other than the time associated with the first breaks.
Alternatively, one can use a PSV geometry of the walkabove type, in which the wellbore is not strictly vertical, but deviated, or even horizontal. In this case, several locations are chosen for the surface seismic sources, arranged in sequence directly above each receiver. In this configuration, a walk-above type PSV survey aims to imitate the geometry of a vertical well and a zero offset PSV by combining the data collected while the seismic sources on the surface are located directly above each corresponding receiver location in the wellbore.
Consequently, we have been looking for a long time for a means of improving the estimation of the speed of a formation by means of unexploited PSV data associated with the rising wave field and time windows other than the time associated with the first breaks. .
BRIEF DESCRIPTION OF THE DIFFERENT VIEWS OF THE DRAWINGS In order to facilitate the understanding of the embodiments described and the other advantages of these, reference is now made to the following description, offered in conjunction with the accompanying drawings, in which :
FIG. 1 is a diagram illustrating an example of a vertical seismic profiling system (PSV) according to the described embodiments;
FIG. 2 is a diagram illustrating an example of a PSV system deployed in association with a wellbore and based on a zero offset configuration according to the embodiments described;
FIG. 2A is a diagram illustrating an example of a PSV system deployed in association with a wellbore and based on a walk-above type configuration according to the embodiments described;
FIG. 3 is a block diagram showing an example of an information processing system according to embodiments of the present invention;
FIG. 4A is a diagram illustrating an example of logging while drillinglYANO environment;
FIG. 4B is a diagram illustrating an example of a wireline log environment;
FIG. 5 is a plot of PSVZO data associated with the falling wave field according to embodiments of the present invention;
FIG. 6 is an enlargement of an area extracted from the layout presented in FIG. 5 and a corresponding appearance according to embodiments of the present invention;
FIG. 7 is a semblance using a linear motion analysis by summation of oblique traces of a sliding window of traces of the PSVZO data associated with the falling wave field represented in FIG. 5;
FIG. 8 is a plot of PSVZO data associated with the rising wave field according to embodiments of the present invention;
FIG. 9 is a semblance using a linear motion analysis by summation of oblique traces of a sliding window of traces of the PSVZO data associated with the ascending wave field represented in FIG. 8;
FIG. 10 is a plot of PSVZO data associated with the ascending wave field transformed into bidirectional time according to embodiments of the present invention;
FIG. 11 is a semblance using a linear motion analysis by summation of oblique traces of a sliding window of traces of the PSVZO data associated with the transformed ascending wave field represented in FIG. 10;
FIG. 12 is a flow diagram illustrating the operations of a workflow using an NSGA II algorithm according to embodiments of the present invention; and FIG. 13 is a flow diagram illustrating the operations of a method implemented by a processing system of a PSV system according to embodiments of the present invention.
DETAILED DESCRIPTION OF THE DISCLOSED EMBODIMENTS] 0025] The following discussion is offered to enable those skilled in the art to make and use the invention. Various possible modifications will be obvious to those skilled in the art and the general principles described herein can be applied to embodiments and applications other than those presented below in detail, without departing from the spirit or the scope of the embodiments as defined here. The embodiments described are not limited to the particular embodiments shown; on the other hand, they should be given the widest possible scope taking into account the principles and characteristics described here.
As used here, the terms "couple" or "coupled" refer either to an indirect connection, or to a direct connection. Thus, if a first device couples to a second device, this connection can be made via a direct connection, or via an indirect electrical or mechanical connection via other devices and connections. The term "upward" as used herein means along a drill string or a borehole from a distal end to the surface; the term "descending" as used herein means along the drill string or the borehole, from the surface to the distal end.
It will be understood that the expression "oil well drilling equipment" is not intended to limit the use of the equipment and methods thus described for drilling an oil well. This term also includes the drilling of natural gas wells or oil wells in general. In addition, these wells can be used for production, monitoring or injection as part of the extraction of hydrocarbons or other materials present in the subsoil. They can also be geothermal wells, designed to provide a source of thermal energy instead of hydrocarbons.
As will be understood by those skilled in the art, the aspects of the present disclosure can be implemented in the form of a system, method or computer program. Accordingly, aspects of this disclosure may take the form of a fully hardware embodiment, a fully software embodiment (including firmware, resident software, micro-codes, etc.) or a mode of implementation combining software and hardware aspects, all of which can generally be designated by the terms "circuit", "module" or "system". In addition, aspects of this disclosure may take the form of a computer program incorporated into one or more computer-readable medium (s) and incorporating a computer-readable programming code.
For the purposes of this disclosure, an information processing system may include any device or set of devices for calculating, classifying, processing, transmitting, receiving, retrieving, generating, switching, storing, displaying, setting evidence, detect, record, reproduce, manage or use information, intelligence or data, of any kind, for commercial, scientific or control purposes, in particular. Examples of common computer systems, environments, and / or configurations usable with the information processing system include, but are not limited to, personal computer systems, server computer systems, thin clients, heavy clients, devices or laptops, multiprocessor systems, microprocessor systems, set-top boxes, consumer programmable electronic devices, network PCs, minicomputers, central computer systems and distributed data processing environments comprising any of the above systems or devices or any other suitable device of varying size, shape, performance, functionality and price.
The information processing system can include a variety of media readable by a computer system. These supports can be any available support accessible by the information processing system, and include both volatile or non-volatile supports and removable or non-removable supports. The information processing system may include media readable by a computer system in the form of a volatile memory, such as a random access memory (RAM) and / or a cache memory. The information processing system may further comprise other removable / non-removable, volatile / non-volatile computer system storage media, one or more processing resource (s) such as a unit. central processing unit ("CPU") or hardware or software control logic, and / or read-only memory (ROM). Other components of the information processing system may include one or more network port (s) allowing communication with external devices, as well as various input and output ("I / O") devices such as a keyboard, mouse, or video screen.
The information processing system may also include one or more buses for transmitting communications between the various hardware components. A first device can be communicatively coupled to a second device, if it is connected to the second device via a wired or wireless communication network allowing the transmission of information.
The examples of certain embodiments given below aim to facilitate understanding of the present disclosure. The following examples are not to be construed as limiting or defining the scope of the disclosure in any way. The embodiments of the present disclosure and its advantages are easier to understand by reference to FIGS. 1 to 13, in which identical reference numbers are used to designate identical and corresponding parts.
Now concerning the drawings, FIG. 1 shows an example of a PSV 100 system according to the embodiments described. The PSV 100 system can be used to analyze an underground formation by means of a PSV survey, associated for example with a geophysical survey, using oil well drilling equipment. More specifically, the PSV 100 system can be used to estimate a training speed. The PSV 100 system can use both the rising and falling wave fields to estimate the speed of the formation, where more conventional systems use only the falling wave field. The PSV 100 system can simply calculate the average of the speed estimates from the rising and falling wave fields or, alternatively, combine them in more sophisticated ways using an inversion procedure.
In one embodiment, the PSV system 100 can estimate the training speed using PVS data known as zero offset (zero offset VSP data, the terms “PSV zero offset data” and “PSVZO data” being used here interchangeably) obtained for training. The PSV 100 system can estimate the speed of training by processing a top-down portion of the PSVZO data and generating a first set of speed estimates. The PSV system 100 can then process an ascending portion of the PSVZO data and generate a second set of speed estimates. The first and second sets of speed estimates can then be used to estimate a training speed.
A channel is associated with each seismic receiver. A trace corresponds to the PSV data recorded on each activation of a seismic receiver. In one embodiment, the processing of the downlink portion of the PSVZO data may include applying an oblique trace summation analysis to a range of channels associated with the downlink portion of the PSVZO data, and processing of the uplink portion PSVZO data may include applying an oblique trace summation analysis to the range of channels associated with the ascending portion of the PSVZO data. In one embodiment, the application of oblique trace summation analysis to the range of channels associated with the ascending portion of the PSVZO data and to the range of channels associated with the descending portion of the PSVZO data may include the generation of 'a semblance which is a function of the slope and the time difference. The slope is the slope of the arrival times determined for each trace in a ribbon of traces. The time offset is an arrival time at the first channel in the analyzed channel ribbon, which can be a first pause at the level of a first channel associated with the trace ribbon.
The appearance is determined as a function of a summation of the arrival times in the time window of the arrival times for each trace of the strip of traces, the arrival times taking into account the time difference associated with the slope of the trace.
The trace ribbon comprises the PSV data which correspond to a channel ribbon, each trace of the trace ribbon corresponding to a channel of the channel ribbon. The channel ribbon is a channel sub-range of a channel range. A new trace ribbon is obtained each time the channel ribbon gradually moves or slides along the channel range.
As seen in FIG. 1, the PSV system 100 can comprise a processing system 120, one or more seismic receivers 102 connected in a communicating manner to the processing system 120 and one or more seismic sources ) 104 which applies (s) seismic energy to an underground formation near a wellhead, in a configuration known as PSV with zero offset (PSVZO) (also known to those skilled in the art under the name "near zero offset PSV" - near zero offset - the seismic source is actually placed near the wellhead, not at the level of the latter).
Each seismic source 104 (also known as "firing") is a device which generates controlled seismic energy and directs this energy into the underground formation. Seismic source 104 can generate seismic energy in various ways, for example using an explosive device (e.g. dynamite or other explosive charge), an air gun, a generator truck shock waves, a seismic vibrator or other devices capable of generating seismic energy in a controlled manner. Seismic sources 104 can deliver point pulses of seismic energy or perform continuous scanning of seismic energy.
The seismic receiver 102 (geophone, hydrophone or distributed acoustic sensor, for example) is a device used in the acquisition of seismic data which detects the speed of the ground produced by seismic waves and transforms the movement into electrical pulses. Three seismic receivers 102 a to 102 c are represented and collectively designated by the term "seismic receivers 102", without this having the effect of limiting in any way the number of possible seismic receivers. The seismic receiver 102 can detect movement in a variety of ways, for example by means of an analog device (e.g., a spring-mounted magnetic mass moving inside a cable reel, or a cable to optical fiber detecting backscattered laser light) or a micro-electro-mechanical device (MEMS) (e.g., a MEMS device that generates an electrical signal in response to ground movement through a loop of active feedback). The seismic receivers 102 generate PSV data corresponding to the detected movement.
The processing system 120 comprises at least one processor (which is not expressly shown) which communicates with the seismic receivers 102 and the seismic sources 104 to send and receive information (PSV data, in particular) from the seismic receivers 102 and the seismic sources 104 and controlling the operation of the seismic receivers 102 and the seismic sources 104. The different processors of the processing system 120 can perform different tasks related to data collection, data processing and the control of the seismic sources 104 and seismic receivers 102 a to 102 c. These processors can be physically and / or functionally distributed and operate either autonomously or cooperatively.
FIG. 2 shows an example of a physical arrangement of the PSV 100 system based on a zero offset configuration. For a set of PSV data with zero shift (PSVZO) (obtained in an area where the geological structure is plane or in strata), choosing the time of the first pauses of each receiver makes it possible to calculate the speed (or the slowness) of the formation from the slope of these first breaks. We know that in this scenario, the upward reflected wave field operates the same time shifts as the downward wave field during its upward propagation towards the surface of the ground. However, the rising wave field is not currently used to determine the speed of a formation. The embodiments described use both the falling and rising wave fields to estimate the speed (or slowness) of the training.
As shown in FIG. 2, one or more seismic source (s) 104 is / are positioned on the surface 108 of the underground formation 110, while the seismic receivers 102 a to 102 c are positioned inside a wellbore 150. When several seismic sources 104 are used, these are positioned close to each other, so that they can be treated as a single seismic source 104 for analysis purposes.
In some cases, the underground formation 110 can be heterogeneous and include several sets of different media (eg, rock, clay, sand, etc.). The formation 110 may include at least one interface 106 between these different environments. The seismic energy generated by the seismic sources 104 moves through the underground formation 110. Part of this energy is reflected and / or refracted by the constituent elements of the underground formation 110 (eg, reflected by the at least an interface 106). The seismic receivers 102 can detect the reflected and / or refracted seismic / acoustic energy and can generate PSV data representing the detected energy. When the seismic receivers 102 are geophones or hydrophones, each seismic receiver 102 corresponds to a different channel. When the seismic receivers 102 comprise a DAS type optical fiber receiver, this comprises, along its length, a plurality of different channels. Information of a temporal nature (in other words, seismic "traces" of a temporal nature) can be obtained from the PSV data and associated with a channel.
The propagation of seismic energy through a medium and the generation of the seismic traces which result therefrom depend on various factors. For example, the speed of propagation can depend on the properties of the medium, such as its density, its elasticity and its depth below the surface. Thus, the seismic energy directed into the underground formation 110 can propagate differently depending on the composition of the underground formation 110.
The arrival time (or "travel time") of the seismic energy at a receiver 102 can also depend on the location of the seismic sources 104, the seismic receivers 102 and the interfaces 106. In for example, the seismic energy from a single seismic source 104 may be associated with different arrival times at each of the seismic receivers 102 a to 102 c, since each of the seismic receivers 102 a to 102 c is located at a different depth below the surface 108. In another example, the seismic energy from different seismic sources 104 can be associated with different arrival times at each of the seismic receivers 102 a to 102 c, because each seismic source 104 is located at a different point along surface 108.
The seismic traces of each of the seismic receptors 102 a to 102 c can be “migrated” as a function of information relating to the known or expected properties of the underground formation 110. Migration is a process in which each sample of a input seismic trace is mapped to an output image based on an image point in the basement. For example, seismic traces can be migrated by applying a velocity model which describes the behavior of seismic energy through underground formation 110 based on known information or forecasts on the composition of underground formation 110. If the velocity model used for migration is precise, when the seismic traces are migrated, the reflection events in the migrated output data before summation or migrated grouped images (common image gathers / CIG) are correctly aligned, making it possible to obtain an image clear of the underground formation. However, if an incorrect speed model is used, the reflection events of the migrated output data before summing may not be aligned and the image may be blurred or inaccurate.
In the example of an arrangement of FIG. 2, the seismic sources 104 and the seismic receivers 102 a to 102 c are connected in a communicating manner to the processing system 120 via a communication interface (such as the telemetry described below). An example of a communication interface includes, for example, wired connectors and / or wireless transceivers.
The example of arrangement of the PSV 100 system shown in the
FIG. 2 is not necessarily to scale. In general, the components of the PSV 100 system can be placed in various physical geometries to analyze the underground formation. In an example of geometry, the seismic sources 104 are positioned along the surface 108 of the underground formation 110, the seismic receivers 102 a to 102 c are positioned at depths of 1000 m, 1500 m and 3000 m at below the surface 108, respectively, and the interface 106 is situated at a depth of 2,700 m below the surface 108. It will however be understood that the seismic receivers 102 a to 102 c and the interface 106 can be arranged or located at other depths or locations. In this example, the surface 108 of the underground formation 110 is located on the surface of the ground. However, in certain forms of implementation, the surface 108 may be located on the seabed, be arranged below an overburden, or the equivalent.
The embodiments of the present disclosure can be applied to horizontal, vertical, deviated, multilateral wells, with U-shaped tubing connection, at intersection, to bypass (drilling around an object blocked at mid depth, then resumption of drilling lower in the well) or any other form of non-linear drilling well, regardless of the nature of the underground formation. Certain embodiments can be applied for example to a wired drill rod, to spiral tubes (wired or non-wired), to data acquired by logging by wired line or by smooth cable or by logging during drilling / measurement in drilling course (LWD / MWD). Certain embodiments can be applied to underwater and / or deep water wells. The embodiments described below being a particular form of implementation have no limiting effect.
FIG. 2 is subject to change, addition or deletion, without departing from the scope of this disclosure. For example, the PSV 100 system may be associated with wireline logging, DAS PSV or smooth cable logging activities, including before the completion of wellbore 150. In addition, the PSV system 100 may be subject to additions or deletions of components, without departing from the scope of this disclosure.
In FIG. 2A, the wellbore 150 and the seismic sources 104 are deployed according to an alternative PSV geometry, based on a walkabove type configuration. The wellbore 150 is not strictly vertical, but is deviated, even horizontal. Several seismic sources 104 a to 104 c are deployed on the surface 108 and several seismic receivers 102 a to 102 c are deployed in the wellbore 150. The location of each seismic source 104 a to 104 c is chosen so as to be located just above one of the receivers 102 a to 102 c. The location of the seismic source 104 a is chosen so as to be located just above the seismic receiver 102 a; the location of the seismic source 104b is chosen so as to be located just above the seismic receiver 102b; and the location of the seismic source 104 c is chosen so as to be situated just above the seismic receiver 102 c. In this configuration, a walk-above type PSV reading allows to imitate the geometry of a vertical well and a zero offset PSV (PSVZO) by combining the data collected by seismic sources 104 a to 104 c.
FIG. 3 is a block diagram showing an example of a processing system 120 according to embodiments of the present disclosure. The processing system 120 can be configured to receive PSV data from receivers (eg, seismic receivers 102 shown in FIGS. 1 and 2), and analyze the PSV data so as to execute one or more methods of attenuation of noise, evaluation of the quality of the data, migration of the data, analysis by summation of oblique traces, development of appearances, estimation of the speed of a formation and display of images. A part of the processing system 120 can perform processing of the PSV data collected by different drilling and logging systems, even when these drilling and logging systems are positioned at different locations.
The processing system 120 comprises at least one processor 304. The processor 304 may comprise, for example, a microprocessor, a microcontroller, a digital signal processor (DSP), a specific integrated circuit of an application (application specifies integrated circuit / ASIC), or any other digital or analog circuit configured to interpret and / or execute program instructions and / or process data. As illustrated, processor 304 is communicatively coupled to at least one memory 306 and configured to interpret and / or execute program instructions stored in memory 306 and / or read and / or write data stored in memory 306. The program instructions can be included in one or more software module (s) 308, such as the data collection module 316, the data analysis module 318, the model estimation module. speed 320 and the graphical user interface / GUI module 322.
The memory 306 can include any system, device or apparatus configured to contain and / or contain one or more memory module (s); for example, memory 306 can include read-only memory, random access memory, semiconductor memory or disk memory. Each memory module may include any system, device, or apparatus configured to store program instructions and / or data for a period of time (eg, computer readable non-transient media). For example, instructions given by software modules 316, 318, 320 and 322 can be retrieved and stored in memory 306 for execution by processor 304.
In one embodiment of the present disclosure, the data used or generated by the software modules 316, 318, 320 and 322, eg. PSV data received from receivers 102, results of analysis of PSV data, as well as one or more speed model (s) 330, etc., can be stored in database 312 temporarily or long-term term. In certain embodiments, the processing system 120 may further comprise one or more screens or other input / output devices making it possible to display the information processed by the processing system 120, such as graphical presentations of the appearances and of the PSV data.
The processing system 120 may further include at least one communication port 314 allowing communication with external devices, eg. networked devices or peripheral devices (I / O devices such as a keyboard, mouse or even a video screen). The processing system 120 may include a plurality of individual processing systems, which may for example be networked with each other.
In embodiments, the processing system 120 may include different sub-processing systems which execute the data collection module 316 to collect the PSV data generated by the receivers, the data analysis module 318, the speed model estimation module 320 and the GUI module 322. The different sub-processing systems can be coupled in a communicating manner to at least one other of the sub-processing systems, in particular by a wired or wireless communication link thread. For example, a sub-processing system executing the data collection module 316 can be positioned on the surface 108 of the underground formation 110, near the wellbore 150, while one or more sub-system (s) -processing executing the data analysis module 318, the speed model estimation module 320 and the GUI module 322 can be located at one or more location (s) remote from the wellbore 150. Two or most sub-processing systems may have common components, for example processor 304, memory 306, database 312 and / or communication port 314, or integrate their own individual components.
In embodiments, the data received by the data collection module 316 can be simulated PSV data which can be received, for example, from a simulator, from an external data center or from a server. storage hosting a PSV data library.
FIG. 3 may be subject to modification, addition or deletion, without however departing from the scope of this disclosure. For example, FIG. 3 shows a configuration of components of the processing system 120.11 is nevertheless possible to use any suitable configuration of components. For example, the components of the processing system 120 can be implemented either in the form of physical components, or in the form of logical components. Furthermore, in certain embodiments, the functionality associated with the components of the processing system 120 can be implemented in circuits or components for special use. In other embodiments, the functionality associated with the components of the processing system 120 can be implemented in general purpose and configurable circuits or components. For example, the components of the processing system 120 can be implemented by configured computer program instructions.
FIGS. 4A and 4B show examples of oil well drilling equipment and drilling environments with which the described PSV system can be used. FIG. 4A presents an appropriate context for describing the operation of the systems and methods described in an illustrated drilling well (LWD) environment. A drilling platform 402 is equipped with a derrick 404 which supports a lifting machine 406 making it possible to raise and lower a drilling train 408. At the lifting machine 406 is suspended an upper drive 410 which rotates the drill string 408 when it is lowered into a wellhead 412. A drill bit (not shown) which rotates to create the wellbore 150 in the formation 110 can be connected to the lower end of the drill string 408. A bottomhole assembly (BHA) module (not shown) may be provided near the drill bit to collect data.
A pump 416 circulates drilling fluid in a supply pipe 418 to the upper drive 410, inside the drill string 408, through the orifices present in the drill bit, then from back to the surface and finally into a retention basin 424. The drilling fluid transports the cuttings from the wellbore 150 into the basin 424 and helps maintain the integrity of the wellbore 150. The drilling fluid is often called "mud" in the petroleum industry; a distinction is often made between water-based drilling fluids and hydrocarbon-based drilling fluids (depending on the type of solvent used).
Data from seismic receivers 102 can be transmitted using various telemetry tools used in drilling operations. The seismic receivers 102 can be coupled to a telemetry module 428 for transmitting the telemetry signals. These telemetry signals can be transmitted to a receiving device 430 on the surface 108 of the wellbore 150. The receiving device 430 can be integrated into the processing system 120 or be put in communication therewith to supply the telemetric signals to the system. 120. The transmission of telemetry signals can be carried out by one or more devices, such as a downhole receiver, which receives the telemetry signals sent by the telemetry module 428, and / or repeaters. downhole, which receive and then retransmit the telemetry signals for reception by the telemetry device 430 at the surface 108 of the wellbore 150.
For example, the telemetry module 428 may include an acoustic telemetry transmitter which transmits telemetry signals in the form of acoustic vibrations in the cased wall of the drill string 408. The downhole receiver can be coupled to the casing below the upper drive 410 to receive the transmitted telemetry signals. Downhole repeaters may include one or more repeater modules 432 which may be optionally provided along the drill string 408 to receive and retransmit telemetry signals. Other telemetry techniques can be used, including telemetry based on mud pulse transmission, electromagnetic telemetry and cable drilling tube telemetry. In certain embodiments, the telemetry module 428 also stores, or as a variant, the PSV data generated by the seismic receivers 102, which are then recovered when the telemetry module 428 rises to the surface 108 of the wellbore 150.
FIG. 4B presents another context suitable for describing the operation of the systems and methods described, in which a configuration with a wired line is used. Logging operations can then be performed using a wired line logging tool 450, e.g. a detection instrument by probe, suspended by a cable 456. The cable 456 may comprise conductors for supplying the tool 450 with electricity and / or passing communications from the tool 450 to the surface of the wellbore 150. A log portion of the wired line logging tool 450 may be provided with centering arms 452 which center the tool 450 inside the wellbore 150 when the tool 450 is raised. In some embodiments, the seismic receivers 102 may be mounted on the cable 456 and lowered into the wellbore 150. In other embodiments, the receivers 102 may be channels of a fiber recording system DAS type optics.
As in the logging environment during drilling (LWD) shown in FIG. 4A, telemetry can be used to provide the processing system 120 with the data generated by the seismic receivers 102. The seismic receivers 102 can be coupled to the telemetry module 428, so that the telemetry signals can be transmitted, via one or more repetition module (s) 432 and / or a downhole receiver (not shown), from the seismic receivers 102 to the receiving device 430 located on the surface 108 of the wellbore 150.
A logging installation 460 collects the measurements carried out by the wired logging tool 450 and includes calculation facilities 462 which may include a receiving device 430 making it possible to receive the telemetry signals and / or a processing system 120 for processing and storing the PSV data generated by the seismic receivers 102.
In FIG. 5, the plot 500 represents PSVZO data corresponding to a downward wave field (also called downward PSVZO data), obtained during a test of PSV with zero offset (or almost zero) and a PSV system (such as the PSV 100 system) deployed at a wellbore. The horizontal axis (x-axis) of plot 500 represents the channel numbers associated with different receivers (e.g. geophones or hydrophones) or with channels located along a DAS-type fiber optic receiver, and l the vertical axis (ordinate axis) represents samples which can be taken over time, e.g. periodically. The white dotted box 502 indicates a range of PSVZO data to be analyzed, for example by summation as described in more detail below, the range being based on choices of arrival time.
The arrival time choices are the arrival times chosen from the PSVZO data which correspond to the arrival of the first break choices. The first pause choices correspond to the moment when the seismic receiver detects a significant variation in an ambient or threshold sound level. Arrival times can be obtained automatically or manually. In embodiments, arrival times can be extracted using an algorithm, such as a first pause threshold detection algorithm. However, in noisy conditions, the time associated with the choice of the first break can be entered manually as a choice of initial arrival time. The reference number 504 indicates an example of choice of arrival time detected automatically or entered manually. This choice of arrival time 504 can be used as one choice of initial arrival time among others and the white dotted lines 502 represent a range of data around the choices of initial arrival time usable in the estimation process. of the speed of training.
The slope and the positioning of a straight line of arrival times 506 are determined by interpolating several choices of arrival time 504. This slope is called slope of arrival times. The finish time line 506 can be extended one way or the other. Box 502 is formed by an upper line 508 and a lower line 510 which follow the slope of the finish time line 506. In the example shown, the upper line 508 is located slightly above the finish time line 506 and the lower straight line 510 is located below the finish time line 506, the spacing between the lines 510 and 506 being greater than the spacing between the lines 508 and 506. The space between lines 508 and 510 draws a time window. This time window can be selected according to conditions such as the noise / signal ratio. In the presence of minimal noise, a range can correspond to three cycles and can be extended to a full recording length, as in the presence of noisy conditions.
In certain embodiments, instead of a simple first pause threshold detection algorithm, the arrival times can be determined by means of a linear summation process based on the semblance, called trace summation oblique. The oblique trace summation includes the application of a Radon transform to the PSVZO data; it is performed on a sliding strip (also called a “window”) of a range of channels. Each trace of a trace ribbon includes the PSV data associated with each channel of the channel ribbon. Similarity is a consistency statistic that provides a quantitative measure of the similarity of seismic data from different channels; it can be defined, for example, by Equation (1):
S (r, p) = rgsr / ict + g!) ² ) ΜΣίϊ (ΣΪ'Λ ² (τ + ίι)) ffi-artl (1) where 5 is the semblance value, p is the slope value which indicates the slope of the arrival times for the i th trace of a ribbon of traces, t is the time in a time window defined by the interval tl to t2, f is the i th trace of the ribbon, Af is the number of traces in the ribbon, and δ is the observed time offset (i.e. the time offset between the first trace of the ribbon and the current trace z) associated with the linear slope p for trace z, and τ is the time offset associated with the time window tl at t2. The value of τ can be assigned to the time of the first sample in the time window of the first trace in the trace ribbon, in the middle of the time window of the first trace in the trace ribbon or, in another reasonable manner, at the time window of the trace ribbon, τ varies between the top of the trace and the bottom of the trace, while the range of slopes analyzed is chosen from a reasonable range of slowness of the rock formation, ranging for example from -1 000 to 1000 microseconds / meter.
In other words, the semblance S, which is a function of the slope relative to the time offset associated with the time window, is determined by summing, in the time window, the arrival times for each trace of the ribbon of traces, the arrival times taking into account the time difference associated with the linear slope p of the trace.
FIG. 6 presents two plots 600 and 620 illustrating the PSVZO data of FIG. 5 after reprocessing. The plot 600 represents the downward PSVZO data illustrated in FIG. 5, but only for sixteen channels (2000 to 2015), this group of channels being called "ribbon". Plot 600 includes a white dotted box at the location of a sliding window 602 which (like box 502 in plot 500 of FIG. 5) indicates a range of interest around arrival times associated with the choice of first pause of PSVZO data. Line 620 corresponds to a complete oblique trace summation analysis of line 600. The vertical axis of line 620 represents the time shift τ, and the horizontal axis represents the range of slopes p.
Referring to Equation (1), the size of the time interval tl to t2 used can be selected so as to optimize the quality of the input data. This selection can be made, for example, by sliding the sliding window 602 up or down to obtain a portion of large optimal amplitude, indicated in 628. The semblance values S determined according to Equation (1) are color coded using a 624 gray scale, in which the darker shades indicate greater amplitude and consistency. The semblance values S determined according to Equation (1) are plotted as semblance data, indicated in 626.
The area of interest for the analysis shown in plot 620 relates to the area shown in the white dotted box at the location of the sliding window 602, which corresponds to the slope of the first break of Descending PSVZO data for the location of the sliding window 602, corresponding to the time interval il to t2 in Equation (1). The corresponding range of the oblique trace summation analysis represented in the plot 620 subject of the study is indicated by the black dashed box 622. Thus, the oblique trace summation analysis can be carried out for the area d 'interest, rather than for all values of τ (time lags).
The high amplitude part 628, shown in black, of the appearance data 626 reported in the trace 620 represents the best coherence associated with the traces which correspond to the ribbon of channels represented in the trace 600. The maximum value of semblance (represented by the darkest data among those plotted on the plot) of the high amplitude part 628 corresponds to a slope of about 400 micro-seconds / meter, which indicates the best linear movement through the ribbon of channels.
The oblique trace summation analysis is repeated iteratively for each next channel ribbon when the sliding window 602 is moved by incrementing the first channel of the ribbon to the next channel and by sliding the ribbon along the range of channels shown in plot 500 of FIG. 5. In this example, the channel ribbon used in the next iteration would include channels 2001 through 2016.
The quantity of calculations performed and the data produced by the oblique trace summation analysis can be reduced by processing only the descending PSVZO data which correspond to a selected arrival window, such as the first arrival window used in this example. The first arrival window corresponds to a unique set of values per channel ribbon corresponding to a unique value of τ, where τ is associated with the arrival time of the first pause on the trace corresponding to the first channel of the channel ribbon. In embodiments, the calculations can be performed using an arrival window different from the first arrival window (for example, a second, third, fourth window, etc.).
In other words, since the analysis by summation of oblique traces mainly relates to the movement, or the slope, of the first break itself (represented by the box in white dotted lines at the current location of the sliding window 602 in plot 600, and denoted tl to t2 in Equation (1)), the range of application of the oblique trace summation analysis can be limited to the black dashed box 622 in the FIG. 6. It is not necessary to apply the full analysis to all time shifts τ, but only to a window around the event of interest. The high amplitude part 628, represented in black in the trace 620, indicates the best coherence for the traces of the trace 600 and thus the best linear movement on the 16 traces near the time sample 400. We then drag the window sliding 602 along the channel range of trace 500 (FIG. 5), select the following set of traces (eg, channels 2001 to 2016) and repeat the analysis by summation of oblique traces.
In addition, since only one set of coherence values must be obtained for the first arrival window, the analysis by summation of oblique traces can be compressed into a single set of values per location of the sliding window 602 , corresponding to a unique value of τ representing the time of arrival of the first break on the first trace in the sliding window 602. FIG. 7 represents the "compressed" oblique trace summation analysis.
In FIG. 7, the plot 700 represents the appearance, indicated in 702, of the downward PSVZO data represented in the plot 500 of FIG. 5. The semblance 702 was obtained by applying the oblique trace summation analysis only to the first arrival window, thus reducing the calculations and the amount of output data. Plot 700 is also called linear motion analysis by summing oblique traces of a sliding ribbon of traces of descending PSVZO data. The vertical axis of the plot 700 represents the slope (ie the inverse of the speed, also called slowness). The horizontal axis of trace 700 represents the channel number, for example, for the entire range of channel numbers illustrated in trace 500. It is understood that, since there is a reciprocal relationship between speed and slowness, determining or estimating slowness or speed means that the other (speed or slowness) is also determined or estimated by applying this reciprocal relationship.
The solid white line 704 represents the maximum value of the semblance 702 for each channel, the maximum value being reported for the first channel of each ribbon of channels. The maximum semblance values represented by the solid white line 704 are an estimate of the adequacy of the PSVZO data to the slowness values associated with the respective trace ribbons of the slippery trace ribbons tested. Thus, the maximum semblance values give an indication of the adequacy between the slowness values and the PSVZO data included in the trace ribbons tested. A similar type of analysis can be performed for an ascending wave field, an example of which associated data set is presented in FIG. 8.
FIG. 8 is a plot 800 of PSVZO data corresponding to the ascending wave field (also called “ascending PSVZO data”), which can be analyzed by applying the same method as for the descending PSVZO data. As for plot 500 of downward PSVZO data presented in FIG. 5, the abscissa axis of the plot 800 represents the channel numbers and the ordinate axis represents the samples taken over time.
In FIG. 8, the arrival times of the ascending PSVZO data can be determined by oblique trace summation based on the linear summation process defined by Equation (1), in which a sliding ribbon is gradually dragged (also called “window”). ) including a sub-range of channels on a range of channels. The oblique trace summation analysis can be performed for an area of interest associated with a selected time range, resembling the white dashed box at the location of the sliding window 602 (in FIG. 6). It can be seen that the ascending PSVZO data represented in the trace 800 are associated with a higher noise level than the descending PSVZO data in the trace 500 represented in FIG. 5.
FIG. 9 is a plot 900 representing the oblique trace summation analysis of the ascending PSVZO data in FIG. 8, obtained by gradually dragging a window over the range of channels shown along the x-axis. This analysis is also called linear motion analysis by summation of oblique traces. As for plot 700 in FIG. 7, the vertical axis of plot 900 represents the slope (i.e. the inverse of speed, also called slowness), and the horizontal axis represents the channel number, for example, for the whole range of channel numbers represented in plot 800. The appearance, indicated in 902, represents the ascending PSVZO data reported in plot 800 of FIG. 8 and was obtained by applying the oblique trace summation analysis only to the first arrival window. The solid white line 904 represents the maximum value of the semblance for each channel, the maximum value being reported for the first channel of each ribbon of channels. The maximum semblance values represented by the solid white line 904 are an estimate of the adequacy of the PSVZO data reported in FIG. 8 to the slowness values associated with the respective trace ribbons of the slippery trace ribbons tested. Thus, the maximum semblance values give an indication of the adequacy between the values of slowness (slope) and the PSVZO data included in the trace ribbons tested. It can be seen that the choices of maximum semblance values represented by the white line 904 are associated with a higher noise level than the corresponding values of the PSVZO data descending from FIG. 7.
Referring to FIG. 10, another way to use the falling wave field is to use the first selected pauses on the falling wave field to transform the rising wave field into bidirectional time. FIG. 10 represents a plot 1000 of several ascending wave fields or reflection events 1002 obtained from the descending PSVZO data. Reflection events 1002 were obtained by using the first pauses chosen from the descending PSVZO data to transform (for example, by adding) the ascending PSVZO data into bidirectional time (the time required to make a round trip from / to the surface). These reflection events 1002 in the plot 1000 represent the upward PSVZO data after application of a downward time offset equivalent to the first pause times estimated from the downward PSVZO data. Once the ascending PSVZO data has been transformed into bidirectional time, oblique summation analysis can be performed to identify the residual time offsets that align with the reflection events 1002.
FIG. 11 is a plot 1100 representing the analysis by summation of oblique traces of the ascending wave field converted into bidirectional time, obtained by gradually sliding a window on the range of channels reported along the abscissa axis in FIG. 10. The appearance is indicated in 1102 and results from a linear motion analysis by summation of oblique traces of the sliding window of traces of ascending PSVZO data. As for plot 700 in FIG. 7, the vertical axis of the plot 1100 represents the slope (i.e. the inverse of the speed, also called slowness), and the horizontal axis represents the channel number, for example, for the whole range of channel numbers illustrated in plot 1000 of FIG. 10. The semblance 1102 represents the ascending PSVZO data converted into bidirectional time represented in the plot 1000 of FIG. 10, obtained by applying the oblique trace summation analysis only to the first arrival window. The solid white line 1104 represents the maximum value of the semblance for each channel, the maximum value being reported for the first channel of each ribbon of channels. The maximum semblance values represented by the white line 1104 are an estimate of the adequacy of the PSVZO data reported in FIG. 10 to the slowness values associated with the trace ribbons of the slippery trace ribbons tested. Thus, the maximum semblance values give an indication of the adequacy between the values of slowness (slope) and the PSVZO data included in the trace ribbons tested.
In FIG. 10, the instability (variation) of the choices of maximum values represented by the solid white line 1104 has been reduced significantly due to the alignment of the reflection events 1002 (i.e. when the PSVZO data ascending have been transformed into bidirectional time by means of descending first pause choices). Each of the reflection events aligned on plot 1000 of FIG. 10 can be used to obtain an estimate of the residual time lags. About ten events 1002 or more clearly visible on plot 1000 could be used independently to obtain estimates of the residual time offsets.
The speed (or slowness) of the training for each channel can then be calculated by means of an algorithm or an inversion procedure using any estimate of slowness among those presented above. Optionally, smoothing and / or filtering can be applied to the slowness estimation values associated with the falling part and the rising part of the PSV data before applying the inversion algorithm. For example, the slowness estimates in FIG. 7 which use the maximum likelihood values associated with the first pause choices of the downward PSVZO data can be used in conjunction with the slowness estimates shown in FIG. 9 which use the maximum likelihood values associated with the first pause choices of the ascending PSVZO data. Alternatively, the inversion algorithm or procedure can use the maximum values of semblance in FIG. 7 with the maximum likelihood values associated with the residual choices of the ascending PSVZO data of FIG. 11. Furthermore, the speed of the formation can be calculated by means of complementary slowness estimates based on different selectable time windows of the ascending wave field converted into bidirectional time shown in FIG. 11. Thus, two or more data sets can be used as inputs to the inversion algorithm or procedure.
Most of the inversion algorithms or procedures are based on the well known inverse problem. Traditionally, the opposite problem has been formulated as shown in Equation (2):
G * m = d, (2) where G is a direct response of the soil based on the acquisition geometry, m is a vector of model parameters to be estimated, and d represents the observed data. The classical solution to this inverse problem is given by Equation (3):
m = (G ^/ Gy ^, G ^/ i / ₍₃₎ However, a solution according to the embodiments described uses a more sophisticated inversion process than that provided by Equation (3). '' Equation (4), formulated below, the inversion method described here uses several sets of observed data associated with the descending and ascending PSVZO data, which are used as inputs and which are processed in a manner to match them with synthetic forecast data:
L0
1111110
111 111 0 0 11 111
0 .... 0 0 0 .... 0 10 0 .... 0 • [Sil 5 ₂ 5 ₃Jtj / DZUTI / DZAf / DZ 00011111 1 -S _n - -4t _n / DZ. G * m = d,
(4) where G is a direct modeling operator (also called “matrix G”), m is a vector of current estimated model parameters (also called “vector m”), and d corresponds to the forecast data ( synthetic) (also called "vector d"). The vector m includes the slowness values of each layer in the model. The vector d includes the slope values which are expected to correspond to the slope (which represents the slowness) calculated from the input data, where Δί is the time offset indicating the movement of the event along the channel ribbon and DZ is the distance corresponding to M traces in the ribbon. The number of ones, 1, in each row of the matrix G corresponds to the number of traces (M) in the ribbon. The positioning of the ones in the matrix G corresponds to the range of channels used in each ribbon. In the example presented for Equation (4), each ribbon has six traces for reasons of simplicity; however, the number of traces included in each ribbon is not limited by this example. In the current example, M = 6 and the matrix G is multiplied by 1/6.
Each set of data corresponding to one of the downward PSVZO data choices or upward PSVZO data choices can be processed using Equation (4). As this equation shows, the proposed inversion scheme uses two or more sets of input data to estimate interval slowness (or speed), which provides a higher degree of confidence than that of the methods using a single set of input data (e.g. only input data linked to the downlink PSVZO data). A first set of data from the two or more input data sets includes the slopes (slowness) between receivers or channels for direct downlink P-wave arrivals (downlink PSVZO data) and a second set of data (or other data sets) includes (includes) slopes obtained from the analysis of the reflected energy of ascending P waves within the same arrangement of receivers or channels as that used for incoming waves P direct (bottom-up PSVZO data).
These different sets of input data can be inverted together to obtain a common set of inverted parameters using, for example, a diagram making it possible to reduce a weighted error function with a gradient optimizer, or by setting l inversion as an optimization problem with multiple objectives.
In one embodiment, the reuse of the reduction scheme of a weighted error function can produce a unique inversion solution which would be biased by the choice of the weighting used to combine the errors of the two data sets.
Furthermore, in an embodiment which uses the optimization problem with multiple objectives, one can find solutions which simultaneously reduce the two errors in the two or more input data sets associated with the downward PSVZO data and to ascending PSVZO data, while satisfying certain constraints relating to the model, for example the interval slowness model, as argued in Deb, K.,
Multi-Objective Optimization Using Evolutionary Algorithms: John Wiley and Sons, Inc, Chapter 2, 2001; and Padhi, A., et. al., Multicomponent Pre-Stack Seismic Waveform Inversion in Transversely Isotropie Media Using a Non-Dominated Sorting Genetic Algorithm, Geophys. J. Int., 196, 1600-1618, 2014. For example, a slope data set (slowness) associated with the descending PSVZO data can be noted dl = [ddnl, ddn_2, ddn_n and a slope data set ( slowness) associated with the ascending PSVZO data can be noted d2 = [dupl, dup_2, dup_n . Therefore, the error or mismatch functions can be defined using Equations (5) and (6):
rt y _dn = ^ fddn _i - s _dn _i) ² (5) n
y »/ = ^ dup_i-s upJŸ (6) where s dn and spup are synthetic arrival time slopes generated by a model of interval slowness whose adequacy is evaluated. Applying this inversion scheme produces a set of solutions called Pareto-optimal solutions that reduce the error determined by Equations (5) and (6). If these solutions are plotted on a plot and the axes defined by the functions described in Equations (5) and (6) are the two inadequacies, then the Pareto-optimal solutions form a front having a convex shape when considered from the origin of the coordinate system used. Consequently, these solutions are non-dominant and another choice of inversion solution or optimal interval slowness model based on this suite of solutions may require more consideration of geological constraints, which may be of a qualitative nature.
The optimization problems with multiple objectives can be solved by means of various available algorithms. Examples of solutions are provided in Deb, K., et. al., A Fast and Elitist Multi-Objective Genetic Algorithm: NSGA-II, IEEE Transaction on Evolutionary Computation, 6, No. 2, 181-197, 2002; and Padhi, A., et. al, 2014. The example solutions use a non-dominated sorting genetic algorithm called NSGA IL. The example algorithm starts with a random mother population of size N. This mother population undergoes steps, such as crossbreeding, mutation and selection by tournaments, to produce a daughter population of size N. The combined population of size 2N can then be sorted into different ranks according to the levels of non-dominance. For example, members of rank 1, where rank 1 is the highest rank, are better than all other solutions, but are not better than each other in terms of all mismatches. Rank 2 members are better than all other ranks except rank 1 members, while being non-dominant among themselves. Then members from various ranks, starting with rank 1, are selected to form a new generation of N members. This process continues until a stop criterion is satisfied. In order to obtain a Paretooptimal front uniformly distributed during the tournament selection phase, the NSGA II favors the least populated member of the population when choosing between two members of the same rank.
FIGS. 12 and 13 are flow diagrams which illustrate the implementation of an exemplary embodiment of a method according to the disclosure. Note that the order of operations presented in FIG. 12 and 13 is not imperative, so that in principle, these different operations can be carried out in a different order from that illustrated and / or simultaneously with each other. Likewise, certain operations can be omitted, different operations can be added or replaced, or certain operations or certain groups of operations can be carried out in a separate application according to the embodiments described here. The operations presented in FIG. 12 and 13 can be performed by the processing system 120 shown in FIGS. 2, 3, 4A and 4B. In particular, the processing system 120 can execute one or more of the software modules 308, instructing the processing system 120 to carry out the operations presented in the flow diagram and described in the disclosure.
[00101] Referring now to FIG. 12, this is a flowchart which illustrates an example of a workflow relating to the NSGA IL algorithm. In operation 1202, we generate an initial population size N. In operation 1204, we calculate the objective vectors ( y). In operation 1206, the non-dominated sorting and the stand distance are calculated. In operation 1208, a selection is made by tournaments, e.g. depending on the crossing and the mutation or the stand. In operation 1210, a size N of the daughter population is determined. In operation 1212, the objective vectors (y) of the daughter population are calculated. In operation 1214, it is determined whether a stop criterion has been satisfied. If, in operation 1214, the answer is No, which means that the criteria for stopping has not been satisfied, then in operation 1216 the mother and daughter populations are combined into a population of size 2N . In operation 1218, non-dominated sorting is carried out on the combined population and the settlement distances are calculated. In operation 1220, N new members are chosen from the combined population to pass to the next generation, after which we continue to apply this process in operation 1208. If, in operation 1214, the answer is Yes, which means that the stop criterion has been satisfied, then in operation 1222 the process is terminated and the solutions obtained are recorded.
[00102] Referring now to FIG. 13, this is a flow diagram representing a method implemented by a processing system, such as the processing system 120 of FIG. 1. In operation 1302, the PSV data are received in response to seismic energy applied to the formation. For example, PSV data can be received from a data collection module such as the data collection module 316 of FIG. 3 or by it. PSV data can be zero offset PSV data. The method shown in the flow diagram can be included in a method implemented by a PSV system such as the PSV system 100 shown in FIG. 1. Although not shown in FIG. 13, the method implemented by the PSV system may include applying the seismic energy by means of a seismic energy source, receiving the PSV data by receivers and recording the PSV data received at the means of a recording device which can be included in the communication system or in communication with it.
In operation 1304, optional, it is possible to receive selected time values which define a time range. The selected time values can be entered, for example, by an operator via a GUI module, such as the GUI module 322 shown in FIG. 3. Operations 1306 and 1310 can be performed by a data analysis module such as the data analysis module 318 shown in FIG. 3. Operations 1308, 1312 and 1314 can be performed by a slowness (or speed) model estimation module 320 shown in FIG. 3.
In operation 1306, a descending part of the PSV data associated with a descending wave field is processed. In operation 1308, a first set of slowness estimation values is generated as a function of the processing of the downlink part of the PSV data. Optionally, operation 1308 can also include smoothing and / or filtering of the slowness estimation values associated with the descending part of the PSV data before applying another processing to these data. In operation 1310, an ascending part of the PSV data is processed, associated with an ascending wave field. In operation 1312, at least a second set of slowness estimation values is generated as a function of the processing of the ascending part of the PSV data. Optionally, operation 1312 can also include smoothing and / or filtering the slowness estimation values associated with the upward portion of the PSV data before applying another processing operation to these data. In operation 1314, a slowness estimate associated with training is determined based on the first set and the at least one second set of slowness estimate values. Consequently, the estimate of slowness (or speed) is determined using top-down and bottom-up PSVZO data.
Operation 1306 may include one or more of operations 1316, 1318 and 1320. In operation 1316, an analysis by summation of oblique traces is applied to the descending part of the PSV data associated with a range of channels. The descending part of the PSV data can be associated with a sliding strip of traces itself associated with the range of channels. The oblique trace summation analysis can be applied to a time range which includes the choice of arrival times for the first break, or to the time range defined by the time values received, the time values being able to be selected so as to that the time range includes the choice of arrival times for breaks different from the first break. In operation 1318, a semblance is generated. The semblance represents a consistency statistic associated with the top-down PSV data.
The semblance can be generated by transforming the traces associated with the respective sub-ranges of the channel range from the recording space as a function of the offset of the receiver with respect to the arrival time detected in a parameter domain of radius as a function of the slope (p) with respect to the determined interception time (tau). The transformation of the traces can comprise the summation of the arrival times associated with the respective sub-ranges of the channel range.
In operation 1320, a maximum value of appearance is determined, which represents the maximum coherence associated with the descending part of the PSV data. The respective maximum likelihood values represent an estimate of the adequacy of the descending parts of the PVS data associated with the respective sub-ranges of the channel range analyzed, which gives an indication of the adequacy between the slope of the respective sub-ranges of the channel range and the PSV data included in the corresponding trace ribbon which is associated with the respective sub-ranges of the channel range.
Operation 1310 may include one or more of operations 1322, 1324, 1326 and 1328. In operation 1322, the upward portion of the PSV data can optionally be transformed into bidirectional time. In operation 1324, an oblique trace summation analysis is applied to the ascending part of the PSV data associated with a range of channels. The ascending part of the PSV data can be associated with a sliding strip of traces associated with the channel range. If the ascending part of the PSV data is transformed into bidirectional time in operation 1322, then the oblique trace summation analysis is applied by means of the transformed ascending PSV data. The oblique trace summation analysis can be applied to a time range which includes the choice of arrival times for the first break, or to the time range defined by the time values received, the time values being able to be selected so as to that the time range includes the choice of arrival times for breaks different from the first break.
In operation 1326, a semblance is generated, which represents a consistency statistic linked to the ascending PSV data. In operation 1328, a maximum value of semblance is determined, which represents an estimate of the adequacy of the ascending part of the PVS data associated with the respective sub-ranges of the range of channels analyzed. Thus, the maximum value of semblance gives an indication of the adequacy between the slope of the respective sub-ranges of the channel range and the PSV data included in the corresponding trace ribbon which is associated with the respective sub-ranges of the range of canals.
Operation 1314 can include operation 1330. In operation 1330, the estimate of the speed associated with the training can be determined by applying a statistical function to the values of slowness associated with the maximum value of semblance, for example by calculating the average of the slowness values or by obtaining an average of the values, and the equivalent. However, the estimate of the slowness associated with the training can, as a variant, be determined by applying an inversion procedure to the slowness values calculated from the maximum semblance values associated with the ascending part of the PSV data and the slowness values calculated from the maximum semblance values associated with the falling part of the PSV data. The inversion procedure may include solving a multi-objective optimization problem. The multi-objective optimization problem can be solved using a genetic non-dominated sorting algorithm (NSGA).
Consequently, the system and the methods described make it possible to estimate the slowness or the speeds associated with a training. A method includes receiving PSV data in response to seismic energy applied to the formation, processing a downlink portion of the PSV data associated with a falling wave field, generating a first set of estimate values as a function of the processing of the falling part of the PSV data, the first set of estimation values estimating at least one value among the slowness and the speed, the processing of an ascending part of the PSV data associated with an ascending wave field , generating a second set of estimate values as a function of processing the ascending portion of the PSV data, the second set of estimate values estimating at least one value from slowness and speed, and obtaining d an estimate of at least one value from the speed and slowness associated with training based on the first and second sets of estimate values.
In embodiments, the processing of the downlink portion of the PSV data may include the application of an oblique trace summation analysis to the downlink portion of the PSV data associated with a range of channels. In some embodiments, processing the upward portion of the PSV data may include applying an oblique trace summation analysis to the upward portion of the PSV data associated with a range of channels.
In addition, in embodiments, the application of the oblique trace summation analysis to at least one of the descending and ascending parts of the PSV data associated with the range of channels can include the generation of a semblance which is a function of the slope and the time difference of a plurality of trace ribbons. Each trace ribbon includes PSV data associated with a corresponding channel ribbon gradually slid along the channel range, wherein the slope of one of the trace ribbons is an arrival time slope for each trace of the ribbon of traces and the time offset of the trace ribbon is an arrival time at a first channel in the channel ribbon. The appearance may be determined as a function of a summation of the arrival times in the time window for each trace of the ribbon of traces, the arrival times taking into account the time offset associated with the slope of the trace of the ribbons.
In embodiments, the method further comprises determining a maximum semblance value of the semblance. The maximum semblance value represents the maximum coherence associated with each of the ascending and descending parts of the PSV data and further gives an indication of the adequacy between the slope of the trace ribbon and the PSV data included in the trace ribbon. In some embodiments, estimating at least one of the speed and slowness associated with the training may include applying a statistical function based on the maximum likelihood value associated with the ascending portion of the PSV data. and on the maximum value of semblance associated with the descending part of the PSV data. In other embodiments, the estimation of at least one value among the speed and the slowness associated with the training may include the application of a slowness inversion procedure corresponding to the maximum values of semblance associated with the training. ascending part of the PSV data and the slowness corresponding to the maximum values of semblance associated with the descending part of the PSV data. In embodiments, the application of the inversion procedure may include solving a multi-objective optimization problem.
In addition, in embodiments, the resolution of the optimization problem with multiple objectives can include the use of a genetic algorithm of non-dominated sorting (NSGA). In embodiments, at least one of the descending and ascending parts of the processed PSV data can be associated with a time included in a time range surrounding the choices of arrival time of a first pause following a predetermined time threshold. Furthermore, in embodiments, the method may also include receiving times that define a time range, wherein the time range includes choices of arrival times of another break than the first break, and at least one of the descending and ascending parts of the processed PSV data comprises PSV data associated with the pause defined by the time range. Furthermore, in embodiments, the upward portion of the PSV data can be transformed into bidirectional time. In embodiments, the PSV data may be included in zero offset PSV data. In embodiments, the method may further include applying seismic energy and recording PSV data.
There is provided a PSV system which comprises at least one seismic energy source applying seismic energy to a formation which is the subject of a PSV survey, at least one receiver defining a plurality of channels and arranged under a formation surface for generating PSV data in response to the detection of seismic energy associated with the applied seismic energy and a processing system. The processing system includes at least one processor and a memory coupled to the processor. The memory stores programmable instructions which, when executed by the processor, instruct the processor to receive vertical seismic profiling (PSV) data in response to the seismic energy applied to the formation, process a portion descending PSV data associated with a falling wave field, generating a first set of estimate values based on the processing of the descending part of the PSV data, the first set of estimate values estimating at least one of the slowness and speed, processing an ascending part of the PSV data associated with an ascending wave field, generating a second set of estimation values as a function of the processing of the ascending part of the PSV data, the second set of estimation values estimating at least one value from slowness and speed, determine an estimate of at least one speed and slowness associated with training as a function first and second sets of speed estimation values.
In embodiments, the processing of at least one of the descending and ascending parts of the PSV data can comprise the application of an oblique trace summation analysis to the PSV data associated with the associated channel range to each of the corresponding descending and ascending parts of the PSV data.
In embodiments, applying oblique trace summation analysis to at least one of the top and bottom portions of the PSV data associated with the range of channels may include generation of a semblance. The generation of the semblance can comprise the transformation of the traces associated with the respective ranges of the channel range from the recording space as a function of the offset of the receiver with respect to the arrival time detected in a range of radius parameter as a function the slope (p) with respect to the interception time (tau) of a plurality of trace ribbons. Each trace ribbon can include PSV data associated with a corresponding channel ribbon gradually slid along the channel range, wherein the slope of one of the trace ribbons is an arrival time slope for each trace of the trace ribbon and the time offset of the trace ribbon is an arrival time at a first channel in the channel ribbon.
In embodiments, the programmable instructions, when executed by the processor, further instruct the processor to determine a maximum semblance value of the semblance, the maximum semblance value representing the consistency maximum associated with each of the ascending and descending parts of the PSV data, giving an indication of the adequacy between the slope of the trace ribbon and the PSV data included in the trace ribbon.
A computer system includes a processor and a memory coupled to the processor, in which the memory stores programmable instructions. When the processor executes the programmable instructions, the processor is instructed to receive vertical seismic profiling (PSV) data in response to seismic energy applied to the formation, processing a descending part of the PSV data associated with a descending wave field , generate a first set of estimation values as a function of the processing of the descending part of the PSV data, the first set of estimation values estimating at least one value among the slowness and the speed, process an ascending part of the associated PSV data to an ascending wave field, generating a second set of estimation values as a function of the processing of the ascending part of the PSV data, the second set of estimation values estimating at least one value from slowness and speed, and determine an estimate of at least one speed and slowness associated with the training as a function of the first and second sets of v speed estimation alues.
In embodiments, the processing of at least one of the descending and ascending parts of the PSV data can comprise the application of an oblique trace summation analysis to the PSV data associated with the associated channel range. to each of the corresponding descending and ascending parts of the PSV data.
Although we have illustrated and described aspects, forms of implementation and specific applications of this disclosure, it will be understood that it is not limited to the construction and to the particular compositions described here and that various changes, modifications and variations can be deduced from the foregoing descriptions, without departing from the spirit or the scope of the embodiments described, as defined in the appended claims.

权利要求:
Claims (10)
[1" id="c-fr-0001]
1. Method for estimating training speeds associated with training, the method comprising:
receiving (1302) vertical seismic profiling (PSV) data in response to seismic energy applied to the formation;
processing (1306) a descending part of the PSV data associated with a descending wave field;
generating (1308) a first set of estimate values as a function of processing the downlink portion of the PSV data, the first set of estimate values estimating at least one value of slowness or speed;
processing (1310) an ascending part of the PSV data associated with an ascending wave field;
generating (1312) a second set of estimate values based on the processing of the bottom-up portion of the PSV data; and determining (1314) an estimate of at least one of the slowness or speed associated with the training based on the first and second sets of estimate values.
[2" id="c-fr-0002]
2. Method according to claim 1, in which the processing of the downward part of the PSV data comprises the application (1316) of an analysis by oblique trace summation to the downward part of the PSV data associated with a range of channels, and / or in which the ascending part of PSV data is transformed into bidirectional time, and / or in which the processing of the ascending part of PSV data comprises the application (1322) of an analysis by summation of oblique traces to the ascending part PSV data associated with the channel range.
[3" id="c-fr-0003]
The method of claim 2, wherein applying the oblique trace summation analysis to at least one of the downward portion of the PSV data and the upward portion of the PSV data associated with the channel range comprises generating (1318; 1324) a semblance which is a function of the slope and time offset of a plurality of trace ribbons, each trace ribbon including PSV data associated with a corresponding channel ribbon gradually slid along the range of channels, the slope of one of the trace ribbons being an arrival time slope for each trace of the trace ribbon and the time offset of the trace ribbon being an arrival time at a first channel in the channel ribbon, in which the appearance is determined on the basis of a summation of the arrival times in the time window for each trace of the trace ribbon, the arrival times taking account of the associated time offset cied to the slope of the trace of the ribbons.
[4" id="c-fr-0004]
4. The method of claim 3, further comprising determining (1320; 1326) a maximum semblance value of the semblance, the maximum semblance value representing the maximum coherence associated with each of the ascending and descending parts of the PSV data which gives an indication of the adequacy between the slope of the trace ribbon and the PSV data included in the trace ribbon.
[5" id="c-fr-0005]
5. Method according to claim 4, in which the determination of the estimate of at least one of the speed and the slowness associated with the training comprises:
applying (1330) a statistical function based on the maximum value of semblance associated with the ascending part of the PSV data and on the maximum value of semblance associated with the descending part of the PSV data, or the application of a procedure of inversion to the maximum value of semblance associated with the ascending part of the PSV data and to the maximum value of semblance associated with the descending part of the PSV data, and / or in which the application of the inversion procedure comprises the resolution of a multi-objective optimization problem, and / or in which the resolution of the multi-objective optimization problem includes the use of a non-dominated sorting genetic algorithm.
[6" id="c-fr-0006]
6. Method according to claim i, in which at least one of the descending and ascending parts of the PSV data which is processed is associated with a time included in a time window which surrounds the choice of arrival times of a first break. according to a predetermined time threshold, and / or wherein the method further comprises receiving (1304) times that define a time range, wherein the time range includes choices of arrival times of a different reflection event d a first pause, and at least one of the descending and ascending parts of the PSV data which is processed includes PSV data associated with the reflection event defined by the time range.
[7" id="c-fr-0007]
7. The method of claim 1, wherein the PSV data includes at least some of the PSV data with almost zero offset and PSV data of the walk-above type.
[8" id="c-fr-0008]
8. The method of claim 1, further comprising: applying seismic energy to the formation; and recording of PSV data.
[9" id="c-fr-0009]
9. Vertical seismic profiling system (PSV) (100) using a method according to any one of claims 1 to 8, the PSV system comprising:
at least one seismic energy source (104) applying seismic energy to a formation subject to a PSV survey;
at least one receiver (102) defining a plurality of channels disposed beneath a surface (108) of the formation (110) for generating PSV data in response to the detection of seismic energy associated with the applied seismic energy; and a processing system (120) comprising: at least one processor (304); and a memory (306) coupled to the processor, wherein the memory stores programmable instructions which, when executed by the processor, instruct the processor to carry out the method according to any of claims 1 at 8.
[10" id="c-fr-0010]
10. Computer system using a method according to any one of claims 1 to 8, the computer system comprising:
a processor;
a memory coupled to the processor, in which the memory stores programmable instructions which, when executed by the processor, give the latter instructions for implementing the method according to any one of claims 1 to 8.

类似技术:

---

公开号 | 公开日 | 专利标题

---

FR3069930A1|2019-02-08|ESTIMATING THE SPEED OF TRAINING BY VERTICAL SEISMIC PROFILING

---

EP2253970B1|2014-01-22|Method for imaging a target area of the subsoil using walkaway data

---

US10571584B2|2020-02-25|Global inversion based estimation of anisotropy parameters for orthorhombic media

---

AU2013392659B2|2017-06-15|Methods and systems for seismic data analysis using a tilted transversely isotropic | model

---

CA2964893C|2019-04-02|Structure tensor constrained tomographic velocity analysis

---

FR3060174A1|2018-06-15|WAVELET ESTIMATION FOR A QUADRIDIMENSIONAL CHARACTERIZATION OF SUBSURFACE PROPERTIES BASED ON DYNAMIC SIMULATION

---

EP2802903A1|2014-11-19|Processing collected survey data

---

CN112424646A|2021-02-26|Seismic data interpretation system

---

FR3061237A1|2018-06-29|ESTIMATE OF THE SLAVE OF QUASI-STATIC STONELEY

---

FR3049355A1|2017-09-29|

---

FR3085053A1|2020-02-21|PREDICTION OF OIL TANK BEHAVIOR USING A PROXY FLOW MODEL TECHNICAL AREA

---

FR3077142A1|2019-07-26|OPTIMIZING A REFERENCE LENGTH FOR SIGNAL PRESERVATION AND PROCESSING A REFERENCE LENGTH FOR DISTRIBUTED VIBRATION DETECTION PURPOSES

---

FR3069929A1|2019-02-08|PREDICTION IN FRONT OF TREPAN USING VERTICAL SEISMIC PROFILE AND GLOBAL INVERSION

---

US10281606B2|2019-05-07|Creating 3C distributed acoustic sensing data

---

US20180059274A1|2018-03-01|Post-critical reflection muting in seismic migration

---

FR3035723A1|2016-11-04|ANISOTROPIC PARAMETER ESTIMATION FROM AUTONOMOUS VSP DATA USING DIFFERENTIAL EVOLUTION

---

US20210270983A1|2021-09-02|Cascaded machine-learning workflow for salt seismic interpretation

---

FR3032532A1|2016-08-12|

---

FR3048782A1|2017-09-15|INTERNAL SEISMIC DATA GENERATOR

---

US20210215841A1|2021-07-15|Bandwith Extension of Geophysical Data

---

US20210318458A1|2021-10-14|Methodology for Enhancing Properties of Geophysical Data with Deep Learning Networks

---

FR3081056A1|2019-11-15|SYSTEM AND METHOD FOR THE APPLICATION OF ELASTIC PROPRIETARY CONSTRAINTS TO THE MODELING OF THE SUBTERRANEAN SURFACE PETROLEASTIC RESERVOIR

---

US10942285B2|2021-03-09|Horizon-based splitting intensity inversion for anisotropic characterization of a target volume

---

US20220018983A1|2022-01-20|Method and device for determining sonic slowness

---

US20210311221A1|2021-10-07|Data-driven domain conversion using machine learning techniques

同族专利:

---

公开号 | 公开日

---

NO20200003A1|2020-01-02|

---

US20200241159A1|2020-07-30|

---

WO2019027466A1|2019-02-07|

引用文献:

---

公开号 | 申请日 | 公开日 | 申请人 | 专利标题

---

US4870580A|1983-12-30|1989-09-26|Schlumberger Technology Corporation|Compressional/shear wave separation in vertical seismic profiling|

---

WO2016106104A1|2014-12-25|2016-06-30|Schlumberger Canada Limited|Seismic sensing and depth estimation of a target reflector|

---

US4802146A|1985-05-23|1989-01-31|Mobil Oil Corporation|Method for moveout correction and stacking velocity estimation of offset VSP data|

---

US4794573A|1988-02-11|1988-12-27|Conoco Inc.|Process for separating upgoing and downgoing events on vertical seismic profiles|

---

US8665667B2|2008-11-08|2014-03-04|1474559 Alberta Ltd.|Vertical seismic profiling velocity estimation method|

---

US9448313B2|2012-02-06|2016-09-20|Ion Geophysical Corporation|Integrated passive and active seismic surveying using multiple arrays|

---

US10073184B2|2012-02-06|2018-09-11|Ion Geophysical Corporation|Sensor system of buried seismic array|

---

US9797243B2|2013-08-22|2017-10-24|Halliburton Energy Services, Inc.|Geophysical prospecting by processing vertical seismic profiles using downward continuation|

---

US20150300161A1|2014-04-22|2015-10-22|Schlumberger Technology Corporation|Down Hole Subsurface Wave System with Drill String Wave Discrimination and Method of Using Same|US11199084B2|2016-04-07|2021-12-14|Bp Exploration Operating Company Limited|Detecting downhole events using acoustic frequency domain features|

---

US20200182047A1|2016-04-07|2020-06-11|Bp Exploration Operating Company Limited|Detecting Downhole Sand Ingress Locations|

---

EA038373B1|2017-03-31|2021-08-17|Бп Эксплорейшн Оперейтинг Компани Лимитед|Well and overburden monitoring using distributed acoustic sensors|

---

EP3673148B1|2017-08-23|2021-10-06|BP Exploration Operating Company Limited|Detecting downhole sand ingress locations|

---

WO2021073740A1|2019-10-17|2021-04-22|Lytt Limited|Inflow detection using dts features|

---

WO2021093974A1|2019-11-15|2021-05-20|Lytt Limited|Systems and methods for draw down improvements across wellbores|

法律状态:
2019-07-30| PLFP| Fee payment|Year of fee payment: 2 |

---

2020-05-01| PLSC| Search report ready|Effective date: 20200501 |

---

2021-05-14| RX| Complete rejection|Effective date: 20210402 |

优先权:

---

申请号 | 申请日 | 专利标题

---

IBWOUS2017045364|2017-08-03|

---

PCT/US2017/045364|WO2019027466A1|2017-08-03|2017-08-03|Vertical seismic profiling formation velocity estimation|

---