U.S. patent application number 15/624403 was filed with the patent office on 2017-12-21 for systems and methods for acquiring seismic data with gradient data.
The applicant listed for this patent is SCHLUMERGER TECHNOLOGY CORPORATION. Invention is credited to Nihed El Allouche, David Fraser Halliday.
Application Number | 20170363756 15/624403 |
Document ID | / |
Family ID | 60659387 |
Filed Date | 2017-12-21 |
United States Patent
Application |
20170363756 |
Kind Code |
A1 |
El Allouche; Nihed ; et
al. |
December 21, 2017 |
SYSTEMS AND METHODS FOR ACQUIRING SEISMIC DATA WITH GRADIENT
DATA
Abstract
A seismic receiver may acquire seismic data reflected from one
or more subterranean features of the Earth. The seismic receiver
may include a housing and a four-component sensor that may measure
four properties of a seismic wavefield. The four-component sensor
may be disposed within the housing. The seismic receiver may also
include a particle motion sensor that may measure a particle motion
of the seismic wavefield in at least one direction. The particle
motion sensor may be disposed within the housing, such that the
four-component sensor and the particle motion sensor are separated
by a distance in a first direction.
Inventors: |
El Allouche; Nihed;
(Cambridge, GB) ; Halliday; David Fraser;
(Cambridge, GB) |
|
Applicant: |
Name |
City |
State |
Country |
Type |
SCHLUMERGER TECHNOLOGY CORPORATION |
Sugar Land |
TX |
US |
|
|
Family ID: |
60659387 |
Appl. No.: |
15/624403 |
Filed: |
June 15, 2017 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
62350310 |
Jun 15, 2016 |
|
|
|
62350371 |
Jun 15, 2016 |
|
|
|
62350349 |
Jun 15, 2016 |
|
|
|
Current U.S.
Class: |
1/1 |
Current CPC
Class: |
G01V 1/162 20130101;
G01V 1/3808 20130101; G01V 1/3852 20130101; G01V 2210/32 20130101;
G01V 2210/1427 20130101; G01V 1/301 20130101; G01V 1/30 20130101;
G01V 2210/47 20130101; G01V 1/28 20130101; G01V 1/282 20130101 |
International
Class: |
G01V 1/16 20060101
G01V001/16; G01V 1/18 20060101 G01V001/18; G01V 1/38 20060101
G01V001/38 |
Claims
1. A seismic receiver configured to acquire seismic data reflected
from one or more subterranean features of the Earth, comprising: a
housing; a four-component sensor configured to measure four
properties of a seismic wavefield, wherein the four-component
sensor is disposed within the housing; and a particle motion sensor
configured to measure a particle motion of the seismic wavefield in
at least one direction, wherein the particle motion sensor is
disposed within the housing, and wherein the four-component sensor
and the particle motion sensor are separated by a distance in a
first direction.
2. The seismic receiver of claim 1, wherein the four-component
sensor is configured to measure the particle motion of the seismic
wavefield in three directions and a pressure of the seismic
wavefield.
3. The seismic receiver of claim 1, wherein the particle motion
sensor is configured to measure a vertical particle motion of the
seismic wavefield.
4. The seismic receiver of claim 1, wherein the particle motion
sensor comprises a geophone or an accelerometer.
5. The seismic receiver of claim 1, comprising a second particle
motion sensor, wherein the four-component sensor is separated from
the second particle motion sensor by the distance in a second
direction that is different from the first direction.
6. The seismic receiver of claim 1, wherein the particle motion
sensor is configured to measure the particle motion of the seismic
wavefield in three directions.
7. The seismic receiver of claim 6, comprising a second particle
motion sensor, wherein the four-component sensor is separated from
the second particle motion sensor by the distance in a second
direction that is orthogonal to the first direction, and wherein
the second particle motion sensor is configured to measure the
particle motion of the seismic wavefield in the three
directions.
8. The seismic receiver of claim 1, wherein the particle motion
sensor is configured to measure a particle velocity of the seismic
wavefield or a particle acceleration of the seismic wavefield.
9. The seismic receiver of claim 1, wherein a first portion of the
housing comprising the four-component sensor is configured to be
disposed at a seabed.
10. The seismic receiver of claim 9, wherein a second portion of
the housing comprising the particle motion sensor is configured to
be disposed within the seabed.
11. A seismic receiver configured to acquire seismic data reflected
from one or more subterranean features of the Earth, comprising: a
housing; a four-component sensor configured to measure four
properties of a seismic wavefield at a first location, wherein the
four-component sensor is disposed within the housing; a first
pressure sensor configured to measure a first pressure of the
seismic wavefield at a second location different from the first
location, wherein the first pressure sensor is disposed within the
housing, and wherein the four-component sensor and the first
pressure sensor are separated by a distance in a first direction;
and a second pressure sensor configured to measure a second
pressure of the seismic wavefield at a third location different
from the second location.
12. The seismic receiver of claim 11, wherein the four-component
sensor is separated from the second pressure sensor by the distance
in a second direction that is different from the first
direction.
13. The seismic receiver of claim 12, wherein the first direction
is orthogonal to the second direction.
14. The seismic receiver of claim 11, wherein the four-component
sensor comprises a particle motion sensor configured to measure
particle motion of the seismic wavefield in three directions and a
third pressure sensor.
15. The seismic receiver of claim 11, wherein a first portion of
the housing comprising the four-component sensor is configured to
be disposed within a seabed.
16. The seismic receiver of claim 15, wherein a second portion of
the housing comprising the first pressure sensor is configured to
be disposed at the seabed.
17. The seismic receiver of claim 16, wherein the first pressure
sensor is part of a second four-component sensor configured to
measure the first pressure and a particle motion of the seismic
wavefield at the seabed.
18. A seismic receiver configured to acquire seismic data reflected
from one or more subterranean features of the Earth, comprising: a
housing; a four-component sensor configured to measure a first
particle motion of a seismic wavefield in three directions and a
pressure of the seismic wavefield, wherein the four-component
sensor is disposed within the housing; a first particle motion
sensor configured to measure a second particle motion of the
seismic wavefield in at least one of the three directions, wherein
the first particle motion sensor is disposed within the housing,
and wherein the four-component sensor and the first particle motion
sensor are separated by a first distance; and a second particle
motion sensor configured to measure a third particle motion of the
seismic wavefield in the at least one of the three directions,
wherein the second particle motion sensor is disposed within the
housing, and wherein the four-component sensor and the third
particle motion sensor are separated by a second distance.
19. The seismic receiver of claim 18, wherein the four-component
sensor is configured to be disposed at a seabed.
20. The seismic receiver of claim 19, wherein the first particle
motion sensor is configured to be disposed within the seabed.
Description
CROSS-REFERENCE TO RELATED APPLICATION
[0001] This application claims priority to and the benefit of U.S.
Provisional Application No. 62/350,310, entitled "A METHOD OF
ACQUIRING SEISMIC DATA," filed Jun. 15, 2016, which is hereby
incorporated by reference in its entirety for all purposes. This
application also claims priority to and the benefit of U.S.
Provisional Application No. 62/350,371, entitled "A METHOD OF
ACQUIRING AND PROCESSING SEISMIC DATA," filed Jun. 15, 2016, which
is hereby incorporated by reference in its entirety for all
purposes. In addition, this application claims priority to and the
benefit of U.S. Provisional Application No. 62/350,349, entitled "A
METHOD OF RECONSTRUCTING A DATASET," filed Jun. 15, 2016, which is
also hereby incorporated by reference in its entirety for all
purposes.
[0002] This application is related to International Publication No.
WO 2016/179060, International Publication No. WO 2016/179206, and
International Publication No. WO 2015/168130, each of which is
hereby incorporated by reference in its entirety for all purposes.
This application is also related to U.S. Provisional Patent
Application No. 62/261,934, which is hereby incorporated by
reference in its entirety for all purposes.
BACKGROUND
[0003] The present disclosure relates generally to systems and
methods for processing seismic data acquired by one or more seismic
receivers to identify locations of hydrocarbon formations or
deposits within subterranean regions of the earth. Seismic
exploration involves surveying subterranean geological formations
for hydrocarbon deposits. A survey typically involves deploying
seismic source(s) and seismic receivers at predetermined locations
at or near the surface of the Earth. The sources generate seismic
waves, which propagate into the subterranean geological formations
creating pressure changes and vibrations along the way. Changes in
elastic properties of a geological formation scatter the seismic
waves, changing the direction of propagation and other properties
of the seismic waves. Part of the energy emitted by the sources is
reflected back from the geological formations toward the surface
and reaches the seismic receivers. Some seismic receivers are
sensitive to pressure changes (e.g. hydrophones), others to
particle motion (e.g. geophones), and surveys may deploy only one
type of receiver or both.
[0004] In response to the detected seismic events, the receivers
generate electrical signals to produce seismic data. Analysis of
the seismic data can be processed to indicate the presence or
absence of probable locations of hydrocarbon deposits.
Additionally, seismic sources and receivers may be used to monitor
hydrocarbon production from a subterranean reservoir and/or other
fluid flow within the reservoir.
[0005] Some surveys are known as "marine" surveys because they are
conducted in marine environments, which may include saltwater
environments, fresh water environments, and brackish water
environments. In one type of marine survey, called a "towed-array"
survey, an array of seismic receiver-containing streamers is towed
behind a survey vessel which also tows one or more seismic sources.
A possible alternative, or addition, to the use of towed streamers
is the use of ocean bottom cables or ocean bottom nodes which
contain seismic receivers. Unlike streamers, these lay on the sea
bed and do not move during recording of seismic data. In such a
survey the seismic sources may be towed by a vessel. There are also
survey procedures in which the seismic sources are stationary (e.g.
attached to a moored buoy).
[0006] Regardless of whether the receivers and seismic source(s)
are moving or stationary, the received data can incorporate effects
resulting from the methodology used to generate the seismic waves
which penetrate into the subterranean and/or undersea geological
formation. These source-side acquisition effects include signatures
of the seismic sources, radiation patterns, residual shot noise,
data irregularity, sparse data sampling, effects from the use of
more than one seismic source, effects from motion of the seismic
source (e.g., if it is moving while data is being acquired), the
effect of the water surface above the seismic source(s), and the
like. As such, the seismic data collected at the receivers (i.e.
the seismic measurements made by the receivers) may be processed to
remove or reduce artifacts which do not correspond to features of
the geological formations that are being surveyed.
BRIEF DESCRIPTION OF THE DRAWINGS
[0007] The present disclosure is best understood from the following
detailed description when read with reference to the accompanying
drawings. Like characters represent like parts throughout the
drawings. It is emphasized that, in accordance with standard
practice in the industry, various features are not drawn to scale.
In fact, the dimensions of various features may be arbitrarily
increased or reduced for clarity of discussion.
[0008] FIG. 1 illustrates a schematic view of a seismic marine
survey arrangement that includes a seismic vibrator array and ocean
bottom receivers, in accordance with one or more embodiments of the
present disclosure;
[0009] FIG. 2 illustrates a block diagram of a computer system that
may perform one or more of the seismic data processing on seismic
data acquired via the seismic marine survey arrangement of FIG. 1,
in accordance with one or more embodiments of the present
disclosure;
[0010] FIG. 3 illustrates a schematic diagram of seismic survey
geometry having one seismic source and multiple seismic receivers,
in accordance with one or more embodiments of the present
disclosure;
[0011] FIG. 4 illustrates example seismic wavefields of a common
source gather that may be acquired via a seismic survey geometry of
FIG. 3, in accordance with one or more embodiments of the present
disclosure;
[0012] FIG. 5 illustrates a diagram of a cost function that may be
associated with the common source gather of FIG. 4, in accordance
with one or more embodiments of the present disclosure;
[0013] FIG. 6 illustrates a schematic example of seismic wavefields
of a common-source gather that may be acquired via a seismic survey
geometry of FIG. 3 without noise components, in accordance with one
or more embodiments of the present disclosure;
[0014] FIG. 7 illustrates a schematic diagram of seismic survey
geometry having multiple seismic sources and one seismic receiver,
in accordance with one or more embodiments of the present
disclosure;
[0015] FIG. 8 illustrates example seismic wavefields of a
common-receiver gather that may be acquired via a seismic survey
geometry of FIG. 7, in accordance with one or more embodiments of
the present disclosure;
[0016] FIG. 9 illustrates a diagram of a cost function that may be
associated with the common receiver gather of FIG. 8, in accordance
with one or more embodiments of the present disclosure;
[0017] FIG. 10 illustrates a flow chart of a method for
simultaneously reconstructing source side data and removing noise
from seismic data acquired via the seismic marine survey
arrangement of FIG. 1, in accordance with one or more embodiments
of the present disclosure;
[0018] FIG. 11 illustrates a schematic survey geometry that may
correspond to the seismic data processed according to the method of
FIG. 10, in accordance with one or more embodiments of the present
disclosure;
[0019] FIG. 12 illustrates a time slice of seismic wavefields
acquired via the seismic marine survey arrangement of FIG. 1, in
accordance with one or more embodiments of the present
disclosure;
[0020] FIG. 13 illustrates a diagram of a unified source-receiver
cost function that may be associated with the time slice of seismic
wavefields of FIG. 12, in accordance with one or more embodiments
of the present disclosure;
[0021] FIG. 14 illustrates a schematic survey geometry that may
correspond to interpolated seismic data determined according to the
method of FIG. 10, in accordance with one or more embodiments of
the present disclosure;
[0022] FIG. 15 illustrates a time slice of seismic wavefields
acquired via the seismic marine survey arrangement of FIG. 1 and
processed to remove noise according to the method of FIG. 10, in
accordance with one or more embodiments of the present
disclosure;
[0023] FIG. 16 illustrates a flow chart of a method for
simultaneously reconstructing source side data and removing noise
from seismic data acquired via the seismic marine survey
arrangement of FIG. 1, in accordance with one or more embodiments
of the present disclosure;
[0024] FIG. 17 illustrates a flow chart of a method for separating
components from seismic data related to pressure and particle
velocity measurements acquired via the seismic marine survey
arrangement of FIG. 1, in accordance with one or more embodiments
of the present disclosure;
[0025] FIG. 18 illustrates a block diagram of a data flow chart for
separating up-going and down-going seismic wavefields from seismic
data related to pressure and particle velocity measurements
acquired via the seismic marine survey arrangement of FIG. 1, in
accordance with one or more embodiments of the present
disclosure;
[0026] FIG. 19 illustrates another flow chart of a method for
separating components from seismic data related to pressure and
particle velocity measurements acquired via the seismic marine
survey arrangement of FIG. 1, in accordance with one or more
embodiments of the present disclosure;
[0027] FIG. 20 illustrates a block diagram of an embodiment of a
seismic receiver that may be used in the seismic marine survey
arrangement of FIG. 1, in accordance with one or more embodiments
of the present disclosure;
[0028] FIG. 21 illustrates a block diagram of a second embodiment
of a seismic receiver that may be used in the seismic marine survey
arrangement of FIG. 1, in accordance with one or more embodiments
of the present disclosure;
[0029] FIG. 22 illustrates a block diagram of a third embodiment of
a seismic receiver that may be used in the seismic marine survey
arrangement of FIG. 1, in accordance with one or more embodiments
of the present disclosure;
[0030] FIG. 23 illustrates a block diagram of a fourth embodiment
of a seismic receiver that may be used in the seismic marine survey
arrangement of FIG. 1, in accordance with one or more embodiments
of the present disclosure;
[0031] FIG. 24 illustrates a block diagram of a fifth embodiment of
a seismic receiver that may be used in the seismic marine survey
arrangement of FIG. 1, in accordance with one or more embodiments
of the present disclosure;
[0032] FIG. 25 illustrates a flow chart of a method for processing
seismic data acquired via (for example) the seismic receivers of
any of FIGS. 20-24, in accordance with one or more embodiments of
the present disclosure;
[0033] FIG. 26 illustrates a block diagram of a sixth embodiment of
a seismic receiver that may be used in the seismic marine survey
arrangement of FIG. 1, in accordance with one or more embodiments
of the present disclosure;
[0034] FIG. 27 illustrates a block diagram of a seventh embodiment
of a seismic receiver that may be used in the seismic marine survey
arrangement of FIG. 1, in accordance with one or more embodiments
of the present disclosure;
[0035] FIG. 28 illustrates a block diagram of an eighth embodiment
of a seismic receiver that may be used in the seismic marine survey
arrangement of FIG. 1, in accordance with one or more embodiments
of the present disclosure; and
[0036] FIG. 29 illustrates a data flow diagram for processing
seismic data acquired via (for example) the seismic receivers of
any of FIGS. 26-28, in accordance with one or more embodiments of
the present disclosure.
DETAILED DESCRIPTION
[0037] One or more specific embodiments will be described below. In
an effort to provide a concise description of these embodiments,
not all features of an actual implementation are described in the
specification. It should be appreciated that in the development of
any such actual implementation, as in any engineering or design
project, numerous implementation-specific decisions must be made to
achieve the developers' specific goals, such as compliance with
system-related and business-related constraints, which may vary
from one implementation to another. Moreover, it should be
appreciated that such a development effort might be complex and
time consuming, but would nevertheless be a routine undertaking of
design, fabrication, and manufacture for those of ordinary skill
having the benefit of this disclosure.
[0038] Generally, seismic data acquired using marine seismic survey
in shallow water (e.g., less than 300 m depth) differs from seismic
data acquired in deeper water (e.g., greater than 300 m depth). For
example, the seismic data acquired in shallower water depths may
include certain properties such as a Scholte wave and other
coherent noise similar to noise (e.g., ground roll) acquired during
land seismic acquisition.
[0039] For example, a marine survey design may account for a
minimum expected velocity of propagation for seismic (e.g.,
acoustic) wavefields to travel across seismic source and/or
receiver arrays of the marine survey design. As such, in marine
survey designs using ocean bottom nodes (OBN) in deeper water, a
minimum velocity of the propagation of the seismic wavefields
across the source array may correspond to a speed of sound in
water. Using this minimum velocity of propagation, seismic data
acquisition parameters (e.g., distance between seismic receivers)
for the seismic receiver may be determine based on the speed of
sound in water. However, in shallower water, a Scholte wave may be
slower than the speed of sound in water. This slower velocity (at
least at the Scholte wave frequencies) may cause seismic data
processing (e.g., source side reconstruction) over large distances
to be difficult and may produce seismic wavefields with significant
noise components that reduce the ability of the wavefields to
accurately indicate locations of hydrocarbon formations within the
earth.
[0040] In addition to the source side sampling/reconstruction
problem (e.g., removal of effects due to positioning and
interference of seismic sources) described above, marine survey
designs that use ocean bottom nodes arrange seismic receivers on
the seabed with spacing between each seismic receiver of
approximately 300 to 400 m. This amount of distance between each
seismic receiver reduces the effectiveness of seismic data
processing processes that use multiple seismic datasets acquired by
multiple receivers (e.g., array processing). As such, the OBN
seismic data acquired by multiple seismic receivers are processed
individually using data acquired by each respective seismic
receiver.
[0041] With the foregoing in mind, to perform effective array-based
seismic data processing (e.g., noise attenuation) in shallower
water depths, seismic receivers along an Ocean Bottom Cable (OBC)
may be separated by approximately 25 to 50 m, thereby involving an
increased number of seismic receivers as compared to the OBN survey
mentioned above. As such, certain embodiments of the present
disclosure are directed towards performing seismic data processing
on seismic data acquired by seismic receivers disposed at various
distances apart while attenuating noise that may be present in the
seismic data due to the shallow water depths, receiver distances,
and the like.
[0042] Indeed, in some embodiments, a computer system may receive
seismic data acquired via a marine seismic survey and may process
the seismic data using certain algorithms (e.g., Extended
Generalized Matching Pursuit (EGMP)) that employ gradient data
(e.g., pressure gradient, spatial velocity gradient) to identify
wavefield components corresponding to coherent noise in the seismic
data and remove the identified wavefield components from the input
seismic data without performing any interpolation/reconstruction
step.
[0043] In addition, in some embodiments, a computer system may
employ a unified seismic data processing framework for simultaneous
source separation and receiver reconstruction to perform
source-side reconstruction while performing coherent noise
attenuation using a multi-channel method. The combination of the
source-side reconstruction and multichannel (i.e., with gradients)
noise attenuation provide for improved seismic data processing that
enables marine seismic surveys to employ seismic receivers at
larger distances apart, thereby increasing the efficiency of the
marine seismic survey. Additional details with regard to performing
the seismic data processing in accordance with the techniques
described above are provided below with reference to FIGS.
1-19.
[0044] In certain embodiments, different sensors may be
incorporated into seismic receivers to acquire seismic data that
may enable the unified seismic data processing framework described
above to perform various types of seismic data processing. These
sensors, in one example, may include a four-component sensor that
measures particle velocity in three directions along with pressure
and at least one other sensor that measures another property
associated with the location of the seismic receiver. That is, the
at least one other sensor may measure a single particle velocity
component, two particle velocity components, three particle
velocity components, pressure, or any other suitable property. The
acquired data from these seismic receivers may enable the seismic
data processing techniques described herein to be performed more
effectively. Additional details with regard to the seismic
receivers used to acquire seismic data to perform various types of
seismic data processing techniques described herein are provided
below with reference to FIGS. 20-28.
[0045] By way of introduction, FIG. 1 illustrates a schematic
diagram of an example marine seismic survey 10 arrangement that
includes a marine vessel 12 that tows a seismic vibrator array 14
through a body of water 16. The seismic vibrator array 14 may
include seismic vibrators 18 that can be activated in response to
activation signals produced by a controller 20. The controller 20
may provide the activation signals via a link 22 to the seismic
vibrator array 14. In the example of FIG. 1, a series 24 of seismic
receivers 26 (sensors) are deployed on the water bottom 30 (e.g.,
seafloor or seabed). The receivers 26 may be deployed in a cable
(e.g., ocean bottom cable) or nodal (e.g., ocean bottom node) form.
In certain embodiments, the disposition of the seismic receivers 26
may include without limitation directly or indirectly at or in
proximity to the water bottom 30 (e.g., seafloor or seabed), with
or without physical contact to the water bottom 30 (e.g., seafloor
or seabed), coupled to the water bottom 30 (e.g., seafloor or
seabed), and the like.
[0046] The seismic receivers 26 may detect wavefields reflected
from a subsurface structure 28 that is underneath an earth surface
(e.g., under water bottom 30, sea floor, or sea bottom). The
seismic receivers 26 may be capable of receiving reflected seismic
wavefields having at least four properties (e.g., particle
velocities, particle accelerations pressure, and/or other wavefield
properties). In some embodiments, the seismic receivers 26 may be
four-component (4C) seismic receivers that may receive seismic data
having four properties in addition to at least one or more
gradients of the four components. In addition, the seismic
receivers 26 may measure characteristics of wave propagation such
as the ellipticity of the wavefield at the seismic receiver 26. The
subsurface structure 28 can include one or multiple subsurface
elements of interest 32. Source wavefields propagated by the
seismic sources 18 are propagated into the subsurface structure 28.
The subsurface structure 28 reflects a part of the source
wavefields, such that the reflected wavefields are detected by the
seismic receivers 26. Measured data (e.g., seismic data) acquired
by the seismic receivers 26 can be communicated to the controller
20 for storage or for processing. Although FIG. 1 describes data
being communicated via the controller 20, it should be noted that
the controller 20 may include any suitable computing device having
one or more processors.
[0047] The seismic vibrators 18 in the seismic vibrator array 14
can be controlled to either be in-phase or out-of-phase to cause
production of an omnidirectional source wavefield or a source
gradient wavefield, respectively. The controller 20 may send
activation signals to the seismic vibrator array 14 to cause the
seismic vibrator array 14 to produce an omnidirectional source
wavefield in a first shot (e.g., first activation of the seismic
vibrator array 14) and to produce a source gradient wavefield in a
second shot. The seismic vibrator array 14 may also be referred to
as a marine vibrator source or a marine vibrator. In certain
embodiments described herein, the seismic vibrator array 14 may
emit source wavefields with alternating directivity patterns.
Additional details regarding the seismic vibrators 18 discussed
herein may be found in International Publication No. WO
2016/179060, which is incorporated herein by reference above.
Although FIG. 1 provides details with regard to a marine seismic
survey, it should be noted that the processing techniques described
below may be performed, in some embodiments, on land seismic
surveys as well.
[0048] After the seismic data (e.g., reflected seismic wavefields)
is acquired by the seismic receivers 26, the seismic data may be
stored via a storage component, database, or the like. The seismic
data may then undergo seismic data processing to convert the raw
seismic data into wavefield data that may be used to identify
hydrocarbon formations or other geological features within the
subsurface structure 28. In some embodiments, the seismic data
processing may also remove noise components and other artifacts
that may be present in the seismic data.
[0049] With the foregoing in mind, FIG. 2 illustrates a block
diagram of a computer system 40 that may perform various types of
seismic data processing of the present disclosure. It should be
noted that the computer system 40 may also be part of the
controller 20 shown in FIG. 1. In some embodiments, the computer
system 40 may include a seismic vibrator control module 42, which
may be executable on one or multiple processors 44 to control
seismic vibrators of the seismic vibrator array 14 via the
controller 20. The computer system 40 may also include a processing
module 46, which may also be executable on the processor(s) 44 to
perform any of the tasks discussed below. In addition, the
processing module 46 may perform a variety of seismic data
processing techniques, such as crossline reconstruction, in-line
reconstruction, up-down source side wavefield reconstruction,
and/or multi-component imaging, and the like. It should be noted
that the processing module 46 may be provided in a computer system
that is separate from a computer system including the seismic
vibrator control module 42. The processor(s) 44 can be coupled to a
network interface 48 (to allow the computer system 40 to
communicate over a network) and a storage medium (or storage media)
50, to store data and machine-executable instructions.
[0050] The storage medium (or storage media) 50 can be implemented
as one or more non-transitory computer-readable or machine-readable
storage media. The storage media can include different forms of
memory including semiconductor memory devices such as dynamic or
static random access memories (DRAMs or SRAMs), erasable and
programmable read-only memories (EPROMs), electrically erasable and
programmable read-only memories (EEPROMs) and flash memories;
magnetic disks such as fixed, floppy and removable disks; other
magnetic media including tape; optical media such as compact disks
(CDs) or digital video disks (DVDs); or other types of storage
devices. Note that the instructions discussed above can be provided
on one computer-readable or machine-readable storage medium, or
alternatively, can be provided on multiple computer-readable or
machine-readable storage media distributed in a large system having
possibly plural nodes. Such computer-readable or machine-readable
storage medium or media is (are) considered to be part of an
article (or article of manufacture). An article or article of
manufacture can refer to any manufactured single component or
multiple components. The storage medium or media can be located
either in the machine running the machine-readable instructions, or
located at a remote site from which machine-readable instructions
can be downloaded over a network for execution. It should be noted
that the components described above with regard to the computer
system 40 are examples and the computer system 40 may include
additional or fewer components as shown.
[0051] In some embodiments, the seismic data acquired via the
seismic receivers 26 may be stored in a database 52 or other
storage component. The database 52 may be accessible to the
computer system 40 via the network interface 48 or some other
communication protocol. In addition, after the computer system 40
performs the seismic data processing, the resulting seismic data
(e.g., noise attenuated) may be stored in the database 52.
[0052] Keeping the foregoing in mind, the embodiments described
herein are related to combining the seismic data processing
techniques related to seismic data acquired using seismic source
arrays (e.g., seismic vibrator array 14) and seismic receiver
arrays (e.g., series 24 of seismic receivers 26). Prior to
discussing the combined approach for processing seismic data, it
may be useful consider the case of processing seismic data received
via a receiver array deployed on a seabed with a single seismic
source in the water, as illustrated in FIG. 3.
[0053] Referring to FIG. 3, a seismic survey geometry 60
illustrates a single seismic source 62 (e.g., seismic vibrator 18)
disposed within the water 16 and a number of seismic receivers 26
disposed on a water bottom (e.g., sea floor or seabed) 30. In one
embodiment, each seismic receiver 26 of FIG. 3 may include at least
one pressure sensor to measure a pressure value and at least one
additional pressure sensor to measure a horizontal gradient of the
pressure value. In certain embodiments, the seismic receivers 26
may be sampled at a spatial interval of approximately 100 m, which
is beyond the Nyquist interval for some shallow water data.
[0054] After the seismic source 62 is activated and the seismic
receivers 26 acquire the reflected seismic wavefields, the computer
system 40 or other suitable processing device may collect the
reflected seismic wavefields as a common-shot gather. An example
common-shot gather 70 is illustrated in FIG. 4. Referring to FIG.
4, the common-shot gather 70 may include a collection of reflected
seismic wavefields 72. In addition to the reflected seismic
wavefields 72, the common-shot gather 70 may include noise
components 74, such as a Scholte wave and/or the like.
[0055] Based on the common-shot gather 70, the computer system 40
may determine a cost function of a portion of the reflected seismic
wavefields 72. For example, FIG. 5 illustrates a slowness spectrum
(e.g., cost function) for a single frequency of a window 76 of the
common-shot gather 70. The slowness spectrum may be used to
identify which parts of the common-shot gather 70 correspond to
noise and which parts correspond to the desired signal (e.g.,
reflected seismic wavefields). For instance, in FIG. 5, p.sub.1 and
p.sub.2 correspond to cut-off slowness values that may be used to
distinguish between noise (e.g., Scholte wave) and the desired
signal. In other words, coherent noise in the acquired seismic data
can be characterized by a defined range of slowness (e.g., a
slowness range p.sub.1 to p.sub.2), outside of which the reflected
seismic wavefields may include coherent noise.
[0056] In certain embodiments, the slowness spectrum may be based
on a planewave decomposition of the acquired seismic data and may
corresponds to the cost function solved using a multichannel
interpolation by matching pursuit (see, e.g., WO 2015/168130) or
the like.
[0057] Applying the multichannel interpolation by matching pursuit
method to the cost function of FIG. 5 may include iteratively
minimizing at each iteration by: (1) selecting the strongest (e.g.,
largest) component in the cost function (e.g., slowness transform);
(2) computing the amplitude corresponding to this component; (3)
forward modelling the desired output wavefield (e.g., an
interpolated version of the input) using the amplitude and
corresponding plane wave component; (4) using the computed
amplitude to forward model the contribution of the component to the
input; and (5) subtracting this contribution from the input data.
In order to distinguish the correct slowness from the noise
components, the computer system 40 may use the gradient of the
wavefield acquired by the seismic receivers 26, as described by
International Publication No. WO 2015/168130.
[0058] With the foregoing in mind, when attempting to obtain a
de-noised version of the acquired seismic data, the computer system
40 may not compute the desired interpolated output wavefield (e.g.,
step (3)--forward modelling the desired output wavefield), as
provided in the multichannel interpolation by matching pursuit
method. Instead, since the coherent noise in the seismic data can
be characterized by a defined range of slowness (e.g., a slowness
outside the range p.sub.1 to p.sub.2), the computer system 40 may
modify the EGMP workflow detailed above, to model the noise
contribution in the seismic data, rather than interpolating the
desired output wavefield, using the chosen component. That is, if
the chosen component lies within the range of slowness p.sub.1 and
p.sub.2, the computer system 40 may add the modeled contribution is
added to a signal output. Alternatively, if the chosen component
lies outside of the range of slowness p.sub.1 and p.sub.2, the
computer system 40 may add the modeled contribution to a noise
output. As a result, the computer system 40 may generate a noise
estimate and a signal estimate. An example signal estimate is
sketched in the schematic of FIG. 6. As shown in FIG. 6, the noise
components 74 previously included in the common-shot gather 70 of
FIG. 4 have been removed, thereby leaving the signal estimate
(which can also be added back to the residual to preserve any
remaining weak events, if desired).
[0059] In addition to a common-shot gather, the acquired seismic
data may be organized according to a common-receiver gather. For
example, FIG. 7 illustrates an example survey geometry 90 that
corresponds to a common-receiver gather, which includes a number of
seismic sources 62 located in the water 16 and a single seismic
receiver 26 located on the water bottom (e.g., sea floor or seabed)
30. In one embodiment, the seismic sources 62 are marine vibrators
that may emit seismic wavefields with alternating directivity
patterns as described in International Publication No. WO
2015/143189 (hereinafter Halliday 2014), which is incorporated by
reference above. The alternating directivity patterns may enable
multi-channel interpolation on the source side. The multi-channel
interpolation may also include joint processing to deal with other
acquisition effects as described by Halliday 2014. By using the
multi-channel interpolation, the computer system 40 may use seismic
data acquired by the seismic receiver 26 with spacing beyond the
spatial Nyquist interval (e.g., a spacing of 90 m).
[0060] Due to the issues that arise when performing seismic data
acquisition in shallow water described above, a common-receiver
gather 100, as illustrated in FIG. 8, may be similar in appearance
to the common-source gather 70 of FIG. 4. Thus, the cost function
of the common-receiver gather 100, as depicted in FIG. 9, is
similar to the common-source gather 70, as depicted in FIG. 4. In
certain embodiments, the seismic data acquired by the
common-receiver gather may be reconstructed using a similar scheme
described above (where the cost function is iteratively minimized)
to give a dense source output. It should be noted that the computer
system 40 may be used to limit the output of the reconstruction to
exclude unwanted events (e.g., ghost waveforms, replicas, noise)
from the interpolated output.
Joint Source Wavefield Reconstruction and Noise Attenuation
[0061] With the foregoing in mind, the computer system 40 may
process seismic data acquired via a common-source gather and a
common-receiver gather using similar techniques (e.g., matching
pursuit based processing) to remove coherent noise from aliased
receiver data and to reconstruct aliased source data. To remove
coherent noise, the computer system 40 may use seismic data having
receiver gradients. To reconstruct the aliased source data, the
computer system 40 may use seismic data that is acquired via marine
vibrators that emit seismic wavefields with different directivities
or directionalities. In certain embodiments, the computer system
may process seismic data using a combination of both techniques.
That is, the computer system 40 may process seismic data acquired
with marine vibrators emitting wavefields with different
directionalities recorded by, for example, pressure sensors and
pressure gradient sensors, such that the seismic data can be
simultaneously reconstructed on the source side and de-noised on
the receiver side. The resulting seismic data would include seismic
data from a dense grid of reconstructed shot points with coherent
noise removed. As such, the computer system 40 may overcome the
issues that slowly propagating coherent noise may produce on
source-side reconstruction, while extending the range of
applicability of marine vibrator acquisition.
[0062] By way of example, the computer system 40 may process the
seismic data acquired via a common-shot gather and a
common-receiver gather together in a higher-dimensional
"source-receiver space." By working in this higher-dimensional
space (i.e., source-receiver space), the computer system 40 may
compute a cost function (e.g., slowness spectrum) may be computed
as a 2-dimensional or higher-dimensional cost function, rather than
a 1-dimensional cost function. As a result, the cost function in
higher-dimensions can be sparser, thereby making it easier to
identify isolated events. In addition, the problems described above
on both the source and receiver sides can be solved
simultaneously.
[0063] FIG. 10 illustrates an example method 110 for simultaneously
processing seismic data acquired by two or more seismic receivers
26 in response to two or more seismic sources 62 at two or more
locations emitting seismic wavefields that change directivity
patterns from shot to shot. The following description of the method
110 is described as being performed by the computer system 40, but
it should be noted that the method 110 may be performed by any
suitable processing device and any number of receivers, sources,
and locations, which numbers may be the same or different from each
other. In addition, although the method 110 is described as being
performed in a particular order, it should be understood that the
method 110 may be performed in any suitable order.
[0064] Referring now to FIG. 10, at block 112, the computer system
40 may receive seismic data (e.g., reflected seismic wavefields)
acquired by two or more seismic receivers 26 in response to two or
more seismic sources 62, such as the seismic vibrator array 14. The
seismic vibrator array 14 may activate seismic sources 62 that emit
wavefields at two or more locations. FIG. 11 illustrates an example
survey geometry 130 that includes multiple seismic sources 62 and
multiple seismic receivers 26. The survey geometry 130 of FIG. 10
combines the survey geometries depicted in FIGS. 3 and 7. In one or
more embodiments, the seismic sources 62 may be activated such that
the directivity pattern changes from shot to shot.
[0065] For each source activation, the resulting seismic wavefield
may be recorded by two or more seismic receivers 26. Each of the
seismic receiver 26 may record a pressure wavefield and a gradient
of the pressure wavefield in at least one direction. As such, each
seismic receiver 26 may include at least one pressure sensor and at
least one pressure sensor to measure the gradient of a pressure
wavefield in at least one direction. Although the following
discussion may include pressure gradient sensors, it should be
noted that the gradient on the seabed may also be estimated from
other measurements, for example, by taking the finite-difference of
two closely located pressure sensors.
[0066] After receiving the seismic data, at block 114, the computer
system 40 may convert the seismic data into a combined-source
receiver domain. FIG. 12 illustrates an example time slice 132
through the combined source-receiver space. As shown in FIG. 12,
one axis corresponds to source position (x.sub.s) and the other
axis corresponds to the receiver position (x.sub.r). In the
combined source-receiver domain, the different arrivals may appear
as three-dimensional waveforms or as two-dimensional waveforms in a
single time slice. For example, coherent noise 134 of FIG. 12
corresponds to the same noise component 74 illustrated in FIG. 4.
However, by working in this higher-dimensional space, as mentioned
above, the cost function may be computed as a two-dimensional or
higher-dimensional cost function, rather than a one-dimensional
cost function only.
[0067] With this in mind, at block 116, the computer system 40 may
transform the seismic data in the combined source-receiver domain
into a cost function (e.g., slowness transform). By way of example,
FIG. 13 illustrates a cost function 136 that is generated by
transforming the seismic data in the combined source-receiver
domain into a unified source-receiver cost function according to a
slowness transform. The cost function 136 includes axes that
correspond to the receiver slowness (e.g., p.sub.r) and the source
slowness (e.g., p.sub.s).
[0068] At block 118, the computer system 40 may iteratively
identify the strongest (e.g., having highest peak, amplitude, or
the like) component(s) in the cost function. As discussed above,
since the cost function in higher-dimensions is more sparse as
compared to the cost function in lower dimensions, the computer
system 40 may more easily identify isolated events that correspond
to coherent noise more easily. That is, the isolated events that do
not correspond to the reflected seismic wavefields may be more
identifiable as being disparate from the reflected seismic
wavefields.
[0069] After identifying the strongest component(s) of the cost
function, at block 119, the computer system 40 may classify each
identified component as a signal component of the seismic data or
as a noise component of the seismic data. The computer system 40
may classify the identified component based on whether the
associated slowness of the strongest component lies within the
expected slowness range for signal or not.
[0070] If the identified component is noise, the computer system 40
may remove the noise component from input data and add the noise
component to a noise output. Conversely, if the identified
component is part of the signal, the computer system 40 may remove
the signal component from input data and add the signal component
to a signal output. That is, at block 120, the computer system 40
may remove the identified component from the seismic data received
at block 112. In some embodiments, the computer system 40, at each
iteration performed during the block 118, may model the desired
output (e.g., signal component or noise component) based on the
strongest component. The computer system 40 may then subtract the
identified noise component from the input seismic data. As a
result, the source sampling of the modeled signal component
corresponds to a dense grid of seismic sources 62 recorded by a
single seismic receiver 26, as illustrated in FIG. 14.
[0071] In embodiments, the computer system 40 may reconstruct the
seismic data based on source points that correspond to a dense
grid, as illustrated in FIG. 14, using a matching pursuit method or
the like. The matching pursuit method may involve transforming data
into a cost function (transform domain) using a chosen basis
function (e.g., linear basis functions), iteratively selecting the
strongest (e.g., largest) component in the transform domain,
computing an amplitude corresponding to that component, computing
and updating the contribution to the output for that component, and
subtracting the component from the input data. When the strongest
(e.g., largest) component has a slowness that is inside the region
defined as containing a signal, the computer system 40 may classify
the component as part of the desired signal and add the component
to a desired output signal model. Conversely, when the strongest
(e.g., largest) component has a slowness that is outside the region
defined as containing a signal, the computer system 40 may classify
the component as noise and add the component to a desired output
noise model.
[0072] After identifying the noise components, the computer system
40 may remove the noise components from the input seismic data, and
the resulting seismic data may not include the coherent noise 134
(e.g., Scholte wave) previously included in the input seismic data.
For example, FIG. 15 illustrates a time slice 138 similar to the
time slice 132 of FIG. 12 with the coherent noise 134 removed in
light of the processing described with respect to the method 110.
The resulting seismic data is achieved by removing the slowly
propagating coherent noise that occurs during source-side
reconstruction, as discussed above. That is, in the combined
source-receiver domain, the removal of coherent noise can also be
applied to the seismic sources. It should be noted that although
the coherent noise removed from the seismic data is described as
being related to a slowness property, in other embodiments, the
computer system 40 may choose any desired slowness limitation in
the combined cost function.
[0073] In certain embodiments (e.g., see FIG. 10, block 119), when
iteratively selecting the strongest (e.g., largest) component in
the transform domain (e.g., slowness transform), the computer
system 40 may determine whether a slowness of the identified
component lies within the region specified as belonging to a
desired signal, compute a contribution of the desired signal for
the desired densely sampled sources and adding to the output (e.g.,
adding the solution for the current iteration to the solutions from
the previous iteration), compute a contribution of the desired
signal to the input data, and subtract the contribution of the
desired signal from the input seismic data. Alternatively, if the
slowness lies outside of the region specified as belonging to a
desired signal, the computer system 40 may compute a contribution
of the noise component to the input seismic data, subtract the
noise component from the input seismic data, and generate a "noise
output" signal based on the contribution of the noise component. It
should be noted that input data for the above process corresponds
the raw seismic data or a selected window of raw seismic data, and
contribution of the desired signal refers to the "matched" portion
of the input data at the current iteration.
[0074] By using the combined source-receiver domain, the computer
system 40 may produce the same type of output as the two separate
source and receiver methods described above, but in a single
processing step. In this way, the embodiments described herein may
prevent errors from one step from affecting the other, improving
the quality of results, as the data become sparser in higher
dimensions (e.g., wavefield components are more easily
distinguished), and allowing slowness limitations to be applied
simultaneously on both the source and receiver side. In addition,
seismic receivers 26 in OBC and OBN acquisitions may include more
space or distance therebetween, thereby allowing seismic surveys to
be conducted using fewer sensing and computing resources.
[0075] Keeping the foregoing in mind, FIG. 16 illustrates another
method 150 for simultaneously reconstructing source side data and
removing noise from seismic data acquired via the seismic marine
survey arrangement of FIG. 1. The following description of the
method 150 will be discussed as being performed by the computer
system 40, but it should be understood that (like the other methods
of the present disclosure) the method 150 may be performed by any
suitable processing device. In addition, although the method 150 is
described as being performed in a particular order, it should be
noted that (like the other methods of the present disclosure) the
method 150 may be performed in any suitable order.
[0076] Referring now to FIG. 16, at block 152, the computer system
40 may receive seismic data, as discussed above with reference to
block 112 of FIG. 10. At block 154, the computer system may select
a window or portion of the received seismic data. The portion of
seismic data may include two or more time samples from the
reflected seismic wavefields from at least two of the activated
seismic sources 62 and at least two of the seismic receivers
26.
[0077] At block 156, the computer system 40 may generate one or
more basis functions suitable to describe the pressure measurements
and pressure gradient measurements provided in the portion of the
seismic data for each seismic source 62 that contributed to the
data provided in the window of seismic data selected at block 154.
The basis functions may include or be any suitable basis function
that characterizes noise components, signal components, and the
like in a variety of domains (e.g., slowness). For example, the
basis functions may be in the time (e.g., tau-p) domain, may
correspond to parabolic or hyperbolic events, may correspond to
curvelets, wavelets, or any other function that may describe and/or
decompose seismic data.
[0078] After generating the basis functions, the computer system 40
may determine a weight for each basis function to model a portion
of the seismic data that corresponds to the window of seismic data
selected at block 154. In embodiments, the weight for each basis
function may correspond to the amplitude of the identified
component (e.g., peak/strongest).
[0079] At block 160, the computer system 40 may identify criteria
to classify the generated basis functions as signal components or
noise components. The criteria may include a determination as to
whether the identified slowness, as described above, is within an
expected range of the signal. The expected range of the signal may
correspond to user defined range, ranges that typically include
signals, and the like.
[0080] After the basis functions are classified, at block 162, the
computer system 40 may model a signal that corresponds to the
desired seismic data output as a sum of the weighted basis
functions classified as signal components. In the same manner, at
block 164, the computer system 40 may model the noise component as
a sum of weighted basis functions classified as noise
components.
[0081] After modelling the signal component and the noise component
of the received seismic data, at block 166, the computer system 40
may interpolate the signal components for additional pseudo seismic
sources at additional pseudo source locations (e.g., source
reconstruction). That is, the computer system 40 may model signal
components using additional basis functions that correspond to
source locations where there was not a physical source.
[0082] At block 168, the computer system 40 may determine whether
another window of seismic data is available. Alternatively, the
computer system 40 may determine whether additional analyses on
other portions of the seismic data are desired to be performed. If
another window of seismic data is available for analysis, the
computer system 40 may return to block 156 and perform the method
150 for the next window of seismic data. If, however, another
window of seismic data is not available for analysis, the computer
system 40 may proceed to block 170 and display the modelled signal
component, the modelled noise component, the interpolated seismic
data, and the like.
[0083] In addition, the data generated by the method 150 may be
stored in the database 52 or another suitable storage device, such
that the data may be analyzed to identify locations of hydrocarbon
deposits and the like. For example, at block 172, the computer
system 40 may perform subsequent processing on the data generated
by the method 150 to identify the hydrocarbon locations or other
features of interest. The identified locations may then be explored
and operations to extract the deposits may occur.
[0084] By employing the presently disclosed embodiments, the
computer system 40 may more efficiently process seismic data
acquired via ocean bottom node or cable systems, including for
4-dimensional or time-lapse applications. As described above, the
input data for the processing operations described above may
include raw acquired seismic data acquired with a marine vibrator
source (e.g., where the marine vibrator emits alternating
directivity patterns) and sea-bed sensors (e.g., 4C, 4C with
additional gradients of one or more of the four components). The
output may include source data on a dense grid with source-related
acquisition effects removed, and the receiver data on the same grid
as the input data was acquired, but with coherent noise removed. As
a result, coherent noise attenuation can be carried out within the
same process as the source-side reconstruction. In this way, the
applicability of the marine vibrator ocean bottom node acquisition
may be extendable to shallower water without the need to perform
additional processing either before or after the source
reconstruction step.
[0085] Moreover, the embodiments described above allow for the
sampling of the seabed receivers to increase in distance (e.g.,
from 25 m to 100 m) without compromising the ability to attenuate
coherent noise. Thus, the seismic data processing involves less
time to record the source data, while using fewer sensors to cover
the same receiver area.
Wavefield Separation During Source-Side Reconstruction
[0086] In addition to simultaneously reconstructing seismic data on
the source side and de-noising the seismic data on the receiver
side as described above, the computer system 40 may also process
the acquired seismic data to separate components of the seismic
data. For example, up-going wavefields and down-going wavefields
may be separated from the acquired seismic data, pressure
wavefields and shear wavefields may be separated from the acquired
seismic data, multiple free up-going wavefields may be separated
from the acquired seismic data, and the like.
[0087] Before discussing the various embodiments for processing the
seismic data to separate these components, it may be useful to note
that in ocean bottom node (OBN) seismic surveys, due to the
distance between each deployed node (e.g., seismic receiver 26),
the reflected seismic data acquired at each node may be processed
individually. That is, common-source gather-type processing may not
be used. Rather, seismic sources 62 may be distributed on a dense
grid (e.g., either in acquisition, or through
interpolation/reconstruction of sparser acquisition grids), such
that the data acquired by each node in the OBN survey may be
processed accurately. Using some assumptions, the presence of the
source grid may enable the data to be processed using 3-dimensional
data processing algorithms. For example, using this dense grid of
shots, the computer system 40 may analyze the acquired seismic data
using wavefield separation and de-multiple processes, prior to
imaging.
[0088] To process the seismic data acquired at each individual
node, the computer system 40 may receive seismic data from seismic
sensors 24 that may include 4-component (4C) sensors. The 4C
sensors may measure pressure and three components of particle
velocity (e.g., x-direction velocity, y-direction velocity,
z-direction velocity). The properties (e.g., recording components)
measured by the seismic receivers 26 may allow some local data
processing steps to be applied, despite that the seismic data is
acquired from a single location. For example, the weighted
combination of pressure and vertical particle velocity may allow
the measured pressure wavefield to be separated into the up-going
and down-going wavefields. In addition, the weighted combination of
all four recording components can also allow the data to be
separated into pressure (P) wave and shear (S) wave arrivals.
[0089] With the foregoing in mind, the presently disclosed
embodiments may enable the computer system 40 to process the
seismic data acquired by the OBN-type seismic survey by
incorporating the receiver-side processing steps during the part of
the processing flow that solves the problems on the source-side.
This is possible by recognizing that the assumptions involved
(e.g., the sea-bed is flat) in performing up-down separation and
de-multiple processes by up-down deconvolution may involve
decomposing the acquired seismic wavefield into slowness or
wavenumber wavefield components on the source-side. These
wavenumber wavefield components correspond to the source wavefields
used to perform source-side reconstruction. That is, instead of
performing seismic data processing using multi-channel
reconstruction, which employs multiple channels on the source-side
to assist in the reconstruction of source points (or conversely on
the receiver-side to assist in the reconstruction of receiver
points), the presently disclosed embodiments may separate
components of acquired seismic data using data acquired by at least
one seismic receiver 26. Additional details with regard to
separating components of acquired seismic data while attenuating
noise components due to the seismic sources 62 will be discussed
below with reference to FIGS. 17-19.
[0090] By way of example, FIG. 17 illustrates a flow chart of a
method 190 for determining up-going and down-going seismic
wavefields in the seismic data acquired by one or more seismic
receivers 26. Although the method 190 is described below in a
particular order, it should be noted that the method 190 may be
performed in any suitable order. In addition, the method 190 will
be described as being performed by the computer system 40, but any
suitable computing device may perform the following method.
[0091] At block 192, the computer system 40 may receive seismic
data acquired by one or more seismic receivers 26 in response to
seismic sources 62 emitting wavefields from two or more source
locations. In some embodiments, the directivity pattern of the
emitted wavefields may change from shot to shot. In ocean bottom
node acquisition, multiple wavefield components may be acquired via
the seismic data and the seismic receivers 26. The wavefield
components may include pressure and three-components of particle
velocity. By way of example, for each source activation, the one or
more seismic receivers 26 may record the reflected seismic
wavefield, which may include the pressure wavefield and the
vertical particle velocity wavefield.
[0092] In the frequency-wavenumber domain, the pressure, P(.omega.,
k), and vertical particle velocity, V.sub.z(.omega., k), recorded
in the water layer can be related by:
V z ( .omega. , k ) = k z .omega. .rho. P ( .omega. , k ) . ( 1 )
##EQU00001##
where k is the horizontal wavenumber vector (k=k.sub.x for 2D
acquisition, and k=[k.sub.x, k.sub.y] for 3D acquisition), .omega.
is the angular frequency, .rho. is the water density, and k.sub.z
is the vertical wavenumber wavefield component. The directionality
of the vertical particle velocity allows the two recording
components to be combined to separate the data into up- and
down-going wavefields:
P up ( .omega. , k ) = 1 2 [ P ( .omega. , k ) + a ( .omega. )
.omega. .rho. k z V z ( .omega. , k ) ] ( 2 ) P down ( .omega. , k
) = 1 2 [ P ( .omega. , k ) - a ( .omega. ) .omega. .rho. k z V z (
.omega. , k ) ] ( 3 ) ##EQU00002##
This separation may be performed in the frequency-wavenumber or
frequency-slowness domain as the scaling factor
.omega. .rho. k z ##EQU00003##
is dependent on wavenumber and frequency (or slowness). The factor
a(.omega.) is a calibration factor that accounts for differences in
the response of the pressure and particle velocity sensors. In some
embodiments, the calibration factor may be determined from the
acquired seismic data by, in one example, identifying part of the
seismic data where only upgoing waves will be present (e.g.,
between the first arrival at the seabed and the first sea-surface
reflection). In this case, the down-going wavefields should be
exactly zero, so equations (2) and (3) can be solved for the
calibration factor. In the embodiments described herein, however,
the separation of up-going waves and down-going waves may be
determined while performing the source-side reconstruction data
processing operations.
[0093] After receiving the seismic data, at block 194, the computer
system 40 may select a window or portion of the seismic data. As
mentioned above with regard to FIG. 16, the window of the seismic
data may include two or more time samples from the seismic data
associated with at least two of the activated sources as received
by one of the seismic receivers 26.
[0094] After receiving the selection of the window of seismic data,
the computer system may, at block 196, generate one or more basis
functions to describe the pressure and particle velocity
measurements for each seismic source 62 contributing to the data
present in the window of seismic data. In certain embodiments, the
basis function may be defined as follows:
b=exp(ik.sub.nx) (4)
where x is a vector describing spatial coordinates of the seismic
sources 24 within the current analysis window, and k.sub.n is the
wavenumber vector corresponding to the chosen wavefield component
at iteration n.
[0095] At block 198, the computer system 40 may determine a weight
for each basis function to model the window of seismic data, as
described above with respect to block 158.
[0096] After determining the weights for the basis functions, at
block 200, the computer system 40 may apply the weights to the
basis functions associated with the pressure measurements and the
particle velocity measurements. For example, if A.sub.P and A.sub.Z
correspond to the determined weights (e.g., output amplitudes)
determined for the pressure and particle velocity data,
respectively, the weights may be applied to the respective basis
functions at the nth iteration as:
P.sub.n(.omega.,k)=A.sub.Pexp(ik.sub.nx) (5)
V.sub.zn(.omega.,k)=A.sub.Zexp(ik.sub.nx) (6)
[0097] At block 202, the computer system 40 may combine the
weighted basis functions using the physical relationship between
the pressure measurements and the particle velocity measurements.
For example, the weighted pressure measurement and particle
velocity measurement described above in equations (5) and (6) may
be combined with equations (2) and (3) to determine the up-going
and down-going wavefields corresponding to the wavefield component
at the nth iteration according to:
P n up ( .omega. , k ) = 1 2 [ P n ( .omega. , k ) + a ( .omega. )
.omega. .rho. k z V zn ( .omega. , k ) ] ( 7 ) P n down ( .omega. ,
k ) = 1 2 [ P n ( .omega. , k ) - a ( .omega. ) .omega. .rho. k z V
zn ( .omega. , k ) ] ( 8 ) ##EQU00004##
[0098] At block 204, the computer system 40 may determine the
separated up-going and down-going seismic wavefields from the
seismic data according to equations (7) and (8). At block 206, the
computer system 40 may interpolate seismic data for additional
pseudo seismic sources based on the determined up-going and
down-going wavefields. That is, the output dataset may be modelled
using additional basis functions corresponding to source locations
where there was not a physical seismic source 62.
[0099] After interpolating the seismic data, at block 208, the
computer system 40 may determine whether another window of seismic
data is available or desired to be processed. If another window of
seismic data is available, the computer system 40 may return to
block 196 and perform the method 190 again with respect to the new
window of seismic data. Alternatively, if another window of seismic
data is not available, the computer system 40 may proceed to block
210 and display the separated components of the seismic data and/or
the interpolated data. In addition, the computer system may store
the data determined via the method 190 in the database 52 or any
other suitable storage component. The computer system 40 may then
perform subsequent analysis of the separated components of the
seismic data and/or the interpolated data to identify locations of
hydrocarbon deposits or the like.
[0100] By performing the processing operation of the method 190,
the computer system 40 may determine the up-going and down-going
wavefields as part of the source reconstruction algorithm, thereby
forgoing i.e. avoiding the need for a separate processing step.
That is, since the seismic data is decomposed into plane-wave (or
wavenumber) wavefield components during the source-side
reconstruction, the wavefield separation results are achieved with
minimal additional computational cost. A block diagram of the
inputs and outputs of the method 190 is illustrated in FIG. 18 for
reference.
[0101] In the foregoing discussion of the method 190, the seismic
data may include a combination of pressure and vertical particle
velocity measurements, but it should be noted that it is also
possible to use other recording components to do different types of
wavefield processing/separation. That is, the presently disclosed
embodiments with regard to the method 190, as well as the other
systems and methods of the present disclosure, should not be
limited to just these two recording components.
[0102] With the foregoing in mind, FIG. 19 illustrates another
method 220 for determining up-going and down-going seismic
wavefields in the seismic data acquired by one or more seismic
receivers 26. Although the method 220 is described below in a
particular order, it should be noted that the method 220 may be
performed in any suitable order. In addition, the method 220 will
be described as being performed by the computer system 40, but any
suitable computing device may perform the following method.
[0103] Referring now to FIG. 19, the blocks 222 and 224 may
correspond to the blocks 192 and 196, respectively, of FIG. 17. At
block 226, the computer system 40 may determine an amplitude for
the strongest (e.g., largest) wavefield component in the basis
functions for the pressure measurement and the particle velocity
measurement. That is, in some embodiments, for a given frequency,
the computer system 40 may use a specified basis function library
to identify the strongest (e.g., largest) wavefield component in
the data across both the pressure and particle velocity
measurements. The computer system 40 may then compute the
corresponding amplitude of the strongest (e.g., largest) wavefield
component for both measurements.
[0104] Generally, the source-side reconstruction operation may
include determining a wavefield made up of a sum of weighted
plane-wave basis functions, according to the matching pursuit
method or the like. During this operation, the computer system 40
may identify the strongest wavenumber (or slowness) wavefield
component at each frequency, and the amplitude of that wavefield
component may be determined based on a description of the seismic
data that may include without limitation the source ghost, source
motion, source signature effects, and the like. The computer system
40 may then identify the basis function corresponding to the
strongest wavenumber, scale the identified basis function by the
determined amplitude, and add the scaled basis function to the
output wavefield.
[0105] If there will be two datasets that the seismic data is
matched with--the pressure dataset and the particle velocity
dataset--in some embodiments, the computer system 40 may determine
an amplitude of a wavefield component for each strongest wavefield
component in both the pressure and the particle velocity
datasets.
[0106] Referring back to FIG. 19, at block 228, the computer system
40 may determine whether the acquired seismic data falls between
the direct wave and the first sea surface multiple. If the seismic
data does fall between the direct wave and the first sea surface
multiple, the computer system 40 may proceed to block 232 and
assume that the down-going wavefield in the seismic data is zero
and compute the calibration factor a(.omega.) accordingly.
[0107] If, however, the seismic data is after the first sea surface
multiple, the computer system 40 may proceed to block 230 and
compute the average of the calibration factors a(.omega.). In
certain embodiments, calibration factors can be computed for each
seismic receiver 26 in turn, or computed as a sum of the
contributions from neighboring seismic receivers 26. After
determining the calibration factor a(.omega.), the computer system
40 may proceed to block 234 and determine whether the identified
wavefield component exists on the pressure measurement and the
particle velocity measurement based on the calibration factor
a(.omega.). That is, the computer system 40 may use the determined
calibration factor a(.omega.) and the corresponding determined
amplitudes to determine whether the wavefield component exists on
both basis functions that describe the pressure and particle
velocity measurements using equation (1) mentioned above.
[0108] If the computer system 40 determines that the wavefield
component exists on both measurements, the computer system 40 may
proceed to block 236. At block 236, the computer system 40 may use
the determined amplitudes to compute the equivalent up-going and
down-going wavefields of the seismic data. The computer system 40
may then, at block 238, remove the identified wavefield component
from the seismic data received at block 222.
[0109] If, however, the computer system 40 determines that the
wavefield component does not exist on both measurements, the
computer system 40 may proceed to block 242. At block 242, the
computer system 40 may remove the identified wavefield component
from the seismic data received at block 222.
[0110] In certain embodiments, the method 220 may be iteratively
repeated for the next strongest (e.g., largest) wavefield component
of the basis functions that describe the pressure and velocity
measurements of the seismic data.
[0111] In addition to the foregoing discussions related to
processing seismic data, while separating the data into up-going
and down-going pressure fields, the computer system 40 may also
compute a horizontal gradient of the pressure by multiplying the
reconstructed basis function by the optimum wavenumber. That is,
using the relationships described in equations (9) and (10) below,
the computer system 40 may determine the P-wave contributions on
the horizontal and vertical components (v.sub.x.sup.p,
v.sub.z.sup.p) within the same workflow described above.
v x P ( .omega. , p x ) = a ( .omega. ) - k .omega. .rho. P (
.omega. , p x ) + 2 c s 2 k 2 .omega. 2 v x ( .omega. , p x ) ( 9 )
v z P ( .omega. , p x ) = ( 1 - 2 c s 2 k 2 .omega. 2 ) v z (
.omega. , p x ) ( 10 ) ##EQU00005##
where c.sub.s is the shear wave velocity at the seabed.
[0112] The P-wave contributions result from summing the up-going
and down-going wavefields on the horizontal and vertical
components. To determine the calibration factor a(.omega.), the
computer system 40 may identifying a range of times where there are
only up-going wavefields in the seismic data. In some embodiments,
the computer system 40 may split the source-side reconstruction
operation into two parts. The first part may perform the
source-side reconstruction operation in the region where there are
only up-going wavefields, and the second part may include
performing the source-side reconstruction operation in the region
where there are both up-going and down-going wavefields.
[0113] In some embodiments, the computer system 40 may determine
the calibration factor a(.omega.) for the second part from the
first part. In other embodiments, for each analysis window that
should have only up-going wavefields, the computer system 40 may
compute and store the values for the calibration factor a(.omega.)
at each iteration and use the average of this value across all
iterations as the calibration factor a(.omega.) for later windows
where both up- and down-going wavefields are present. As a result,
the computer system may compute the calibration factor a(.omega.)
"on the fly" during the source-side reconstruction operation.
[0114] In addition, prior to performing the up-down separation, in
some embodiments, it may be useful to de-noise or filter certain
noise-modes that may be present in the seismic data. An example of
one of the noise-modes may include "shear on vertical" noise, which
corresponds to leakage of shear energy onto the vertical component
geophone. The computer system 40 may filter this noise node using
data acquired from a horizontal component geophone as a noise
reference for adaptive subtraction. In some embodiments, it may be
possible to attempt to identify noise modes, such as the shear on
vertical noise, that occur on one recording component (e.g.,
vertical particle velocity) and not on another recording component
(e.g., pressure). For example, during source-side reconstruction
operations, the computer system 40 may extract weights or amplitude
values A.sub.P and A.sub.Z for a given slowness wavefield
component. If the calibration factor a(.omega.) is known, the
absolute values of A.sub.P and A.sub.Z should be predictable from
one another, since all terms in equation (1) are known. Thus, if
equation (1) does not hold (within some defined threshold), the
computer system 40 may identify the identified wavefield component
can be identified as noise occurring on either the pressure or
particle velocity recording component. The computer system 40 may
then review or analyze the two recording components independently
to identify which recording component includes the noise. As such,
the computer system 40 may identify receiver-related noise while
performing a source-side reconstruction operation.
[0115] By separating the wavefields in the seismic data during
source-side reconstruction, certain embodiments described herein
save both used disk-space (e.g., memory, storage) and processing
effort. As a result, the time related to processing seismic data
and identifying the locations of hydrocarbon deposits may be
reduced. In addition, the improved processing methods of the
present disclosure described herein may enable certain seismic data
processing operations to be performed locally (e.g., onboard the
recording and/or node recovery/harvesting vessel).
[0116] The embodiments described herein may include identifying
noise that exists on one channel that does not exist on another.
For example, horizontally propagating scattered waves in the water
layer may not be recorded on the vertical geophone recording
component, but will be recorded on the hydrophone. Indeed, shear
wave leakage on the vertical geophone may not be recorded on a
hydrophone. As such, provided that the seismic receivers 26 (e.g.,
sensors) are correctly calibrated, the amplitude of an event on one
recording component may be predicted for another, and hence it may
be possible to determine if an event picked during reconstruction
is an event to be reconstructed or is noise to be discarded. As
such, certain embodiments of the present disclosure may avoid
removing the shear on vertical noise by adaptively subtracting the
horizontal recording component data from the vertical recording
component data. Instead, the presently disclosed techniques may
enable identifying and suppressing noise during reconstruction
which may be a more effective way to this part of the data while
preserving storage space and processing power.
Seismic Receiver Designs
[0117] The foregoing processing techniques have been described as
being performed with seismic receivers 26 having 4C sensors that
may measure pressure and three components of particle velocity
(e.g., x-direction velocity, y-direction velocity, z-direction
velocity). In addition to the 4C sensors, the seismic receivers 26
may include at least one additional sensor that may be used to
determine a gradient measurement. Gradient measurements have been
suggested for land seismic to acquire a noise reference that can be
adaptively subtracted from the vertical component of the wavefield
(Edme et al., 2011) and to combine the gradient with the vertical
component of the wavefield to attenuate spatially aliased noise by
matching pursuit (El Allouche et al., 2015).
[0118] In some instances, the lateral gradients may be estimated by
determining a difference between the vertical component of two
adjacent geophones. Alternatively, gradients may also be
approximated from two horizontal geophones placed at two different
depth levels (Goujon et al., 2013).
[0119] For the marine seismic survey 10, the gradient of the
pressure field may be measured using accelerometers. For example,
pressure data and acceleration data may be combined to separate the
pressure field in up-going and down-going waves using the matching
pursuit approach described above. In some instances, a cluster of
hydrophones and/or geophones placed in the water 16 may be used to
record the pressure gradients. That is, the multicomponent sensor
cluster may measure the gradients of the pressure field in three
orthogonal directions by determining the difference between data
acquired by two hydrophones placed at a distance from each other or
by measuring the acceleration data in the water.
[0120] With the foregoing in mind, certain embodiments of the
present disclosure may include one or more of various node designs
for the seismic receiver 26 that may allow the recording of spatial
gradients of the pressure along with a translational wavefield. By
recording the gradients along with the translational wavefield,
such embodiments enable a range of improved seismic data processing
operations, such as coherent noise attenuation, wavefield
reconstruction, and wavefield separation, as detailed above.
[0121] More specifically, aspects of the present disclosure are
related to extending the seismic receivers 26 employed in an
ocean-bottom node (OBN) that record four components of the
reflected seismic wavefield (e.g., 3C particle velocity and
pressure) at the seabed. The extension may include adding one or
more hydrophones and/or geophones to measure the spatial gradients
of pressure in the water 16 and the particle velocity gradients (or
acceleration) at the water bottom 30 (e.g., seafloor or seabed). By
adding the additional sensor(s) to measure these properties, the
presently disclosed seismic receiver designs may acquire different
data (e.g., particle velocity or acceleration) that may not
correspond to spatial derivatives of the pressure field in the
water 16 due to the seismic receiver's 26 being coupled to the
water bottom 30 (e.g., seafloor or seabed). Indeed, the
incorporation of the additional sensor(s) enable the computer
system 40 to more effectively perform seismic data processing, such
as up-down separation, P-S separation, wavefield reconstruction,
and noise attenuation processing using the matching pursuit
framework in a sparse receiver configuration.
[0122] Keeping the foregoing in mind, FIGS. 20-24 illustrate
various designs in which a seismic receiver 26 may incorporate
different arrangements of sensors. Referring now to FIG. 20, a side
view 250 of one embodiment of the seismic receiver 26 is
illustrated. The seismic receiver 26 may include a 3C
particle-velocity sensor 252 disposed within a housing 254 of the
seismic receiver 26. In addition, the seismic receiver 26 of FIG.
20 may include a pressure sensor 256 disposed on each side of the
seismic receiver 26. The two pressure sensors 256 may be used to
determine the pressure gradient between sides. It should be noted
that each or of the sensors described as being disposed within the
housing 254 may be in fluid contact with the surrounding medium
(e.g., water 16). In some instances, some of the sensors described
herein may be enclosed within the housing 254. In certain
embodiments, the node configuration described in FIG. 20 may be
coupled to the seafloor.
[0123] Although the seismic receiver 26 of FIG. 20 may acquire
pressure gradient data, in certain embodiments, it may be
beneficial to include the additional sensors within the housing 254
of the seismic receiver 26. With this in mind, FIG. 21 illustrates
a side view 260 of another design of the seismic receiver 26 that
may be used in the marine seismic survey 10 described herein. In
one embodiment, the seismic receiver 26 of FIG. 21 may include a 4C
sensor 262 that measures the 3C particle velocity and pressure. The
4C sensor 262 may be disposed on one end or corner of the seismic
receiver 26. In addition to the 4C sensor 262, the seismic receiver
26 may include a 1C particle velocity sensor 264 disposed on an
opposite end of the seismic receiver 26 with respect to the 4C
sensor 262. It should be noted that the sensors described herein
are not limited to be disposed at certain ends or corners of the
seismic receiver 26. Indeed, each of the designs of the seismic
receivers 26 described below include sensors that are positioned,
such that the sensors of a respective seismic receiver 26 include a
certain distance (e.g. less than half a meter) between another one
of the sensors therein. By way of example, in FIG. 21, the 4C
sensor 262 may be disposed a distance (e.g., less than half of a
meter) away from the 1C particle velocity sensor 264. The 1C
particle velocity sensor 264 may measure the particle motion (e.g.,
velocity, acceleration) in one direction. This additional sensor
may be a hydrophone and/or a geophone placed at a fixed distance
266 (e.g., less than half a meter) away from the 4C sensor 262.
Both sensors may thus be located on the same node of an OBN
receiver. The spatial gradients may be estimated by determining a
difference between the data acquired by two hydrophones or
geophones. The seismic receiver designs having one spatial gradient
may be deployed on a rope 268 at a sparse spacing (e.g., greater
than 50 m) and may record the gradient along the receiver line. The
sensors that measure particle motion as described herein may
include a geophone, an accelerometer, or the like.
[0124] In addition to seismic receivers 26 with one spatial
gradient, FIGS. 22-24 illustrate sample seismic receiver designs
with two spatial gradients. As such, additional sensors including
hydrophones, geophones, and/or one to three component geophones may
be enclosed within the housing 254 of the seismic receiver 26.
[0125] For example, FIG. 22 illustrates a top view 270 of a seismic
receiver design that may include the 4C sensor 262 and two pressure
sensors 272 disposed at two different corners of the housing 254 to
provide two horizontal gradients. In the same manner, FIG. 23
illustrates a top view 280 of a seismic receiver design that may
include the 4C sensor 262 and two 1C particle velocity sensors 264
disposed at two different corners of the housing 254 to provide two
horizontal gradients. In another example, FIG. 24 illustrates a top
view 290 of a seismic receiver design that may include the 4C
sensor 262 and two 3C particle velocity sensors 272 disposed at two
different corners of the housing 254 to provide two horizontal
gradients.
[0126] FIGS. 22-24 thus illustrate examples of seismic receiver
nodes with two spatial gradients that can be deployed as autonomous
systems where the gradients are estimated in at least two
orthogonal directions. It should be noted that nodes with pressure
gradients may allow for the estimation of the horizontal
accelerations (ax, ay) and particle velocities (v.sub.x, v.sub.y)
in the water 16 using the equation of motion provided below:
a x w = .differential. t v x w = - 1 p w .differential. x P , a y w
= .differential. t v y w = - 1 p w .differential. y P , ( 12 )
##EQU00006##
Where P corresponds to pressure and .rho..sub.w represents the
water density.
[0127] With this in mind, nodes with particle velocity gradients
may measure the spatial gradient in the seabed according to:
.differential..sub.xv.sub.z(.omega.,p)=-j.omega.p.sub.xv.sub.z,
.differential..sub.yv.sub.z(.omega.,p)=-j.omega.p.sub.yv.sub.z,
.differential..sub.yv.sub.x(.omega.,p)=-j.omega.p.sub.yv.sub.x,
.differential..sub.xv.sub.y(.omega.,p)=-j.omega.p.sub.xv.sub.y,
(13)
where p=(p.sub.x, p.sub.y, q) is the slowness vector.
[0128] By measuring spatial gradients of pressure and particle
velocity on a seismic receiver node coupled to the seabed, the
computer system 40 may perform a variety of types of seismic data
processing for a range of applications. FIG. 25 illustrates a
general process flow diagram of a method 300 for processing seismic
data acquired by one or more types of seismic receivers 26 depicted
in FIGS. 20-24.
[0129] Generally, the method 300 may include using the computer
system 40 or other suitable device to receive seismic data acquired
via one or more seismic receivers 26, such that each seismic
receiver 26 may include a 4C sensor and at least one other sensor
at a horizontal gradient, as described above with reference to
FIGS. 20-24.
[0130] At block 304, the computer system 40 may process the
acquired seismic data with the data acquired by each respective
sensor disposed within or around the respective seismic receiver
26. The seismic data processing may include without limitation
source-side reconstruction, receiver-side reconstruction, noise
attenuation, wavefield separation, interpolation, imaging
operations, and/or the like. The processed seismic data may be
employed to identify hydrocarbon deposits in the earth, such that
hydrocarbon exploration and extraction equipment (e.g., drill,
pumps) may be installed near the location of the hydrocarbon
deposits to produce the hydrocarbons, or other subterranean
feature(s) of interest.
[0131] After processing the seismic data, the computer system 40
may display the processed data or store the processed seismic data
in the database 52, a storage component, or the like. By way of
example, a few of the seismic data processing operations that may
be performed at block 304 are discussed below with respect to a
type of seismic receiver design may be best suited to perform the
respective operation.
Sparse Acquisition for Increased Efficiency
[0132] In shallow water surveys, OBNs are densely deployed in one
direction. In this configuration, the nodes are attached by a rope
(or deployed as part of a cabled system) and placed every 25 to 50
m. In certain embodiments, the seismic receivers of FIG. 20, 22, or
23 may be used to acquire spatial gradients. The data acquired by
these sensors may enable the computer system 40 to reconstruct
source-side seismic data while de-noising the seismic data in
accordance with the following description.
[0133] By acquiring spatial gradients, the seismic receivers 26 may
enable the spacing between the nodes of the OBN system to be
increased. Using the reconstruction by matching pursuit approach
mentioned above, the computer system 40 may represent the pressure
and its spatial gradient as the sum of the same basis functions
(e.g., an exponential basis function) and may be combined to
de-alias the acquired seismic data and reconstruct it on a denser
grid according to:
( P ( x , .omega. ) .differential. x P ( x , .omega. ) ) = ( 1 - j
.omega. p x ) A e j .omega. p x x ( 14 ) ##EQU00007##
[0134] When the spacing between the nodes is increased, the number
of nodes per area decreases, thereby extending the spatial coverage
of the receiver spread. Consequently, the number of receiver patch
deployments can be reduced, which may enable a rolling receiver
spread, rather than a series of patch deployments. Furthermore,
since the required source areal coverage is larger than the
receiver area, the number of redundant/repeated shot points may
also be reduced, thereby significantly increasing the efficiency of
shallow water OBN acquisition.
[0135] While reconstructing the pressure field, the computer system
40 may also reconstruct the horizontal component of particle
velocity/acceleration in the water 16 can also be reconstructed
using equation 12, such that:
A x ( .omega. , p ) = - 1 .rho. w p x P ( 15 ) ##EQU00008##
[0136] This estimated horizontal component of acceleration/particle
velocity in water can be combined with the horizontal component
measured in the water-bottom (i.e., using the OBN) to infer the
medium properties of the seabed.
Up and Down Separation by Matching Pursuit for Sparse
Configuration
[0137] In certain embodiments, the seismic receivers 26 of FIG. 20,
22, or 23 may be used to acquire a horizontal gradient of the
pressure field that can be measured by the sensors within the
respective seismic receivers 26. The data acquired by these sensors
may enable the computer system 40 to estimate separate up-going and
down-going components in the seismic data using the matching
pursuit method in accordance with the following description.
[0138] As discussed above, separating up-going and down-going
wavefields from the seismic data may be part of the processing of
multicomponent seabed data where the up-going pressure field
P.sup.- (reflections from the subsurface) is separated from the
down-going pressure field P.sup.+ (surface-related multiples) by
combining pressure and vertical component measurements in the water
column:
P + = 1 2 P + .rho. w 2 q w v z P - = 1 2 P - .rho. w 2 q w v z (
16 ) ##EQU00009##
where
q w = 1 c w 2 - p x 2 ##EQU00010##
is the vertical slowness, c.sub.w and .rho..sub.w are the wavespeed
and density of the water, respectively. If the computer system 40
receives pressure gradient data from the seismic receivers 26, the
computer system 40 may reconstruct and decompose the pressure field
in a sparser configuration, such that the seismic data is separated
into the up- and down-going pressure wavefields as the seismic data
is being interpolated. In this case, the up and down separation
applied in the shot gather domain, can be formulated as
multichannel input data:
( P .+-. .differential. x P .+-. v z .+-. ) = ( 1 - j .omega. p x
.+-. q w j .omega. .rho. w ) A e j .omega. p x x ( 17 )
##EQU00011##
[0139] These input channels can be used to formulate a cost
function that allows the reconstruction of the pressure field in
the water. The computer system 40 may decompose the seismic data
while reconstructing the seismic data by formulating a cost
function for the up-going pressure field that suppresses the
down-going wavefield (by attributing the corresponding sign of the
vertical slowness to the desired field). Alternatively, the
computer system 40 may reconstruct the pressure and vertical
particle velocity data separately and retrieve the amplitude from
the vertical component. Depending on the sign of the amplitude, the
reconstructed basis function can be identified as up-going or
down-going. In both approaches, the computer system 40 may
calibrate the vertical component of particle velocity. Generally,
the calibration takes into account the difference in coupling as
well as the fact that the particle velocity is measured at the
water bottom 30 (e.g., seafloor or seabed) and not in the water
16.
Coherent Noise Attenuation by Matching Pursuit for Sparse
Configuration
[0140] In certain embodiments, the seismic receivers 26 of FIG. 21,
23, or 24 may be used to acquire a horizontal gradient of the
vertical particle velocity that can be measured by the sensors
within the respective seismic receivers 26. The data acquired by
these sensors may enable the computer system 40 to attenuate
coherent noise in the seismic data using a matching pursuit method
in accordance with the following description.
[0141] In a shallow water environment, the wavefield generated by a
seismic source interacts with the water bottom 30 (e.g., seafloor
or seabed) in the near-field. As a result, Scholte waves (also
known as mud-roll) and guided waves may be excited. Generally, the
seismic data acquired in this type of environment (e.g., the
Persian Gulf and the Red Sea) share many similarities with land
data. Specifically, the low-frequency slowly-propagating Scholte
waves have similar properties to those of ground-rolls present in
land seismic surveys. Therefore, noise attenuation methods based on
gradients developed for sparse land acquisition (where conventional
array-based methods fail) are also be applicable for shallow water
environments.
[0142] For example, when performing noise attenuation by matching
pursuit (El Allouche et al., 2015), the computer system 40 may
separate aliased coherent noise while reconstructing the
multicomponent input data. The separation is based on a wavenumber
filter, where noise is expected to be dominant in a specific
frequency-wavenumber (or slowness) range, which may be outside the
signal cone. As such, the noise attenuation method may keep the
signal in the residual signal and subtract the more energetic noise
from the residual signal. This workflow, as described by El
Allouche et al. (2015), is valid for the existing multicomponent
OBN and OBC systems where there are no available spatial gradients.
However, for the OBN with seismic receiver designs illustrated in
FIGS. 22-24, this approach can be extended to the pressure gradient
(in a similar manner as equation 14) or equivalently to the
vertical particle velocity gradient
( v z ( x , .omega. ) .differential. x v z ( x , .omega. ) P ( x ,
.omega. ) ) = ( C ( .omega. ) - j .omega. p x C ( .omega. ) .+-.
.rho. w j .omega. q w ) A e j .omega. p x x ( 18 ) ##EQU00012##
where C(.omega.) denotes the coupling filter accounting for
differences in coupling between the hydrophone and the
geophone.
[0143] By adding more channels, the computer system 40 may
constrain the problem of de-aliasing. That is, because the coherent
noise is expected to be present on the pressure as well as the
particle velocity (noise exist in the same wavenumber range but
with different phase and amplitude), the computer system 40 may
attenuate the noise on all components even in the case of having
one spatial gradient. This is possible because the computer system
40 can combine one multichannel cost function (e.g. pressure and
pressure gradient) aimed at defining the optimum wavenumber with a
single channel cost function aimed at determining the phase and
amplitude for each component individually (vertical and horizontal
particle velocity). The computer system 40 may identify the noise
on the multichannel cost function and may subsequently subtract the
noise from each component separately using the single channel cost
function.
Coherent Noise Attenuation by Adaptive Subtraction
[0144] In other embodiments, the seismic receivers 26 of FIGS.
20-24 may be used to acquire a horizontal gradient of the vertical
particle velocity or the pressure that can be measured by the
sensors within the respective seismic receivers 26. The data
acquired by these sensors may enable the computer system 40 to
estimate seabed properties and attenuate noise in the seismic data
using adaptive subtraction in accordance with the following
description.
[0145] Using the acquired seismic data, the computer system 40 may
address the coherent noise (e.g., Scholte waves and guided waves)
using noise attenuation by adaptive subtraction. For instance, the
computer system 40 may use the estimates of the horizontal
gradients of the vertical particle velocity or pressure to provide
a scaled version of the same physical quantity. In some
embodiments, the scaling may correspond to the horizontal
wavenumber, as described in equations 13.
[0146] Since the coherent noise described above is slower than
reflections from the subsurface, the computer system 40 may scale
the horizontal gradients of the vertical particle velocity or the
pressure with greater values. As such, the gradient estimate, which
may be dominated mainly by coherent noise, may be used as a noise
reference that can be adaptively subtracted from the pressure or
particle velocity wavefields (Edme et al., 2011). In this way, the
analysis is performed trace-by-trace and therefore is independent
of the spatial sampling.
[0147] As discussed above, in addition to the Scholte wave and
guided waves, another type of coherent noise that may be present on
the vertical component may be known as "shear noise on the
vertical." This noise may generally be attributed to bad coupling
and/or to scattering from small heterogeneities near the seismic
receiver 26 (Paffenholtz et al., 2006). By acquiring the horizontal
gradients of the vertical component in two directions (e.g., OBN
shown in FIG. 23), the computer system 40 may estimate a noise
reference of the slowly propagating shear wave and may adaptively
subtract the noise reference from the vertical component.
Seabed Properties Estimation and P- and S-Wave Separation by
Matching Pursuit
[0148] In other embodiments of the present disclosure, the seismic
receivers 26 of FIGS. 20, 22, and 23 may be used to acquire a
horizontal gradient of the pressure field that can be measured by
the sensors within the respective seismic receivers 26. The data
acquired by these sensors may enable the computer system 40 to
estimate seabed properties and perform P-wave and S-wave separation
in accordance with the following description.
[0149] With the foregoing in mind, it should be noted that elastic
separation of the wavefield into pressure-waves (P-waves) and
shear-waves (S-waves) may not be performed while the seismic data
acquired at the seabed is being processed. As such, the computer
system 40 may assume that the horizontal component contains all of
the S-waves and the vertical component contains all of the P-waves.
A proper elastic separation, however, may involve data regarding
medium parameters at the water-seabed interface.
[0150] Schalkwijk et al. (2008) showed that the spatial gradient of
the pressure field is related to the horizontal component of the
particle velocity according to:
.differential..sub.xP(.omega.,p.sub.x)=-j.omega..rho.(v.sub.x.sup.p-2c.s-
ub.s.sup.2p.sub.x.sup.2v.sub.s) (19)
[0151] Hence, by acquiring the spatial gradient of the pressure
field, the computer system 40 may then estimate the S-wave velocity
(c.sub.s). In embodiments, the P-wave velocity can be estimated by
applying a time window on an event identified on the pressure and
the vertical component data as described by Schalkwijk et al.
(2003). After the seabed medium parameters are estimated, the
computer system 40 may separate the up-going P- and S-wave using
the matching pursuit approach. For each iteration n, the computer
system 40 may thus determine:
P n up ( .omega. , p ) = - p x 2 q p P n ( .omega. , p ) + C (
.omega. ) [ .omega. p k z v z , n ( .omega. , p ) + c z - 2 - 2 p x
2 2 q p v x , n ( .omega. , p ) ] S n up ( .omega. , p ) = - p x 2
q s P n ( .omega. , p ) + C ( .omega. ) .rho. s c s 2 [ v z , n (
.omega. , p ) - c s - 2 - 2 p x 2 2 q s v x , n ( .omega. , p ) ] (
20 ) ##EQU00013##
[0152] P.sub.n(.omega., p), v.sub.x,n(.omega., p) and
v.sub.z,n(.omega., p) are the components of the input wavefield
solved for at the n.sup.th matching pursuit iteration, and
P.sub.n.sup.up and S.sub.n.sup.up denote the result of processing
these components to give the equivalent up-going P-wave and S-wave,
respectively.
Vertical Gradient Measuring Seismic Receivers
[0153] In addition or alternative to the seismic receiver designs
described in FIGS. 20-24, other types of designs may be
incorporated within the seismic receiver 26. For example, FIGS.
26-28 illustrate various designs in which the seismic receiver 26
may include a 4C sensor and at least one additional sensor disposed
to measure a vertical gradient.
[0154] Referring first to FIG. 26, a side view 310 of an example
seismic receiver 26 depicts the 4C sensor 262 disposed at or near
the surface of the water bottom 30 (e.g., seafloor or seabed) and a
2C sensor disposed vertically underneath the 4C sensor 262 by a
distance 314. The seismic receiver 26 of FIG. 26 may thus measure
the vertical gradients, which may be used to estimate the P-wave
and S-wave velocities at the seabed.
[0155] That is, similar to the free-surface on land, the boundary
conditions at the water bottom 30 (e.g., seafloor or seabed) may
include the vanishing of the tangential components of stress
resulting in the following relations between the spatial gradients
of the translational wavefield:
.differential..sub.zv.sub.x=-.differential..sub.xv.sub.z
.differential..sub.zv.sub.y=-.differential..sub.yv.sub.z (21)
[0156] Consequently, the computer system 40 may approximate the
horizontal spatial gradients by taking the difference between two
horizontal translational components buried at different vertical
levels.
[0157] In addition to these applications, the vertical gradients
provide the vertical slowness of the P-waves and S-waves depending
on the type of the wave according to:
.differential..sub.z.differential..sub.x(.omega.,p)=-j.omega.q.sub.pv.su-
b.x,
.differential..sub.zv.sub.x(.omega.,p)=-j.omega.q.sub.sv.sub.x
(22)
[0158] These slownesses are related to the P-wave and S-wave
velocities at the seabed (e.g., the water bottom 30, seafloor, or
seabed) according to:
q p = 1 c p 2 - p x 2 and q p = 1 c s 2 - p x 2 ( 23 )
##EQU00014##
[0159] As such, by measuring the vertical gradients with the
seismic shown in FIG. 26, the computer system 40 may estimate the
P-wave and S-wave velocities at the seabed.
[0160] Referring now to FIG. 27, a side view 320 of another design
for the seismic receiver 26 includes a pressure sensor 272 disposed
at the water bottom 30 (e.g., seafloor or seabed) and the 4C sensor
262 disposed vertically underneath the pressure sensor 272. The
pressure sensor 272 and the 4C sensor 262 may correspond, for
example, to two hydrophones, such that one is in the water 16 and
another one buried under the water bottom 30 (e.g., seafloor or
seabed). By measuring the pressure (P.sub.w) in the water 16 and AT
THE SEABED (P.sub.w), THE COMPUTER SYSTEM 40 MAY RELATE THE
DIVERGENCE (SENSITIVE TO P-WAVES) at the seabed to the pressure
fields in the water 16 according to:
P w = 1 K .gradient. v w = .rho. w c w 2 .gradient. v w P s = .rho.
s ( c p 2 - 4 3 c s 2 ) .gradient. v s ( 24 ) ##EQU00015##
[0161] As a result, the computer system 40 may separate the P-waves
and S-waves from the seismic data acquired at the seabed and may
constrain the medium parameters estimation.
[0162] Referring now to FIG. 28, a side view 320 of another design
for the seismic receiver 26 includes a 4C sensor 262 disposed at
the water bottom 30 and another 4C sensor 262 disposed vertically
underneath the 4C sensor 262 disposed at the water bottom 30 (e.g.,
seafloor or seabed). This configuration is beneficial because the
decoupled geophone in water 16 may detect a signal that is
proportional to the gradients of the pressure field (see equation
12). As such, the computer system 40 may perform many of the
operations discussed above. That is, by recording the pressure and
the three components of the particle velocity/acceleration in water
16 and at the water bottom 30 (e.g., seafloor or seabed), the
computer system 40 may more effectively estimate seabed properties,
the up and down wavefield separation, the P-wave and S-wave
separation, and the like.
[0163] With the foregoing in mind, FIG. 29 illustrates a data flow
diagram 350 that represents example input data and output data of
the various types of systems and methods described herein. Although
the data flow diagram 350 illustrates a number of inputs and a
number of outputs, it should be understood that the various
embodiments of the present disclosure described above may include
one or more of each input and output. In addition, the flow data
diagram 350 is provided as an example diagram and the embodiments
described herein should not be limited to the example data types
provided in FIG. 29.
[0164] Referring now to FIG. 29, the computer system 40 may receive
pressure data 352, horizontal pressure gradient data 354, vertical
particle velocity 356, or any combination of these inputs. The
computer system 40 may then determine a cost function of the
received data at block 358, as described above. Based on the cost
function, the computer system 40 may identify (block 360) optimum
wavenumbers from replicas present in the received data and
distinguish the signal components from the noise components in the
received data, as described above.
[0165] At each iteration, based on the wavenumber, identified
signal, and identified noise, the computer system 40 may output
noise in pressure data 362, noise in pressure gradient data 364,
and noise in particle velocity data 366. These outputs are
subtracted from the pressure data 352, the horizontal pressure
gradient data 354, and the vertical particle velocity data,
respectively.
[0166] In addition, based on the wavenumber, identified signal, and
identified noise, the computer system 40 may output up-going
pressure data 368, down-going pressure data 370, up-going pressure
gradient data 372, down-going pressure gradient data 374, up-going
vertical particle velocity data 376, and down-going vertical
particle velocity data 378. These outputs are subtracted from the
corresponding input data. The process described in the data flow
diagram 350 may be iteratively performed until the seismic data is
determined to accurately represent the desired wavefields.
[0167] Reference throughout this specification to "one embodiment,"
"an embodiment," "embodiments," "some embodiments," or similar
language means that a particular feature, structure, or
characteristic described in connection with the embodiment(s) may
be included in at least one embodiment of the present disclosure.
Appearances of these phrases throughout this specification may, but
do not necessarily, all refer to the same embodiment.
[0168] Although the present disclosure has been described with
respect to specific details, it is not intended that such details
should be regarded as limitations on the scope of the invention,
except to the extent that they are included in the accompanying
claims.
* * * * *