Login| Sign Up| Help| Contact|

Patent Searching and Data


Title:
A METHOD FOR FAST DETERMINATION OF THE LOCATION OF AN ACOUSTIC EMISSION EVENT IN A VTI/TTI MEDIUM
Document Type and Number:
WIPO Patent Application WO/2018/130272
Kind Code:
A1
Abstract:
A method for fast determination of the location of a hypocenter of an acoustic emission event for a Vertical or Tilted Transverse Isotropic formation, each having a symmetry axis, by means of at least three receivers comprising spreading the receivers out perpendicular to a symmetry axis for wave velocities; defining independent source/receiver sagittal planes as planes parallel to said symmetry axis that include the location of the receiver and either the particle movement polarization vector of compressional waves (P waves) or vertical shear waves (Sv waves) generated by the acoustic emission event, or the planes perpendicular to the particle movement polarization vector of a horizontal shear wave (Sh wave) generated by the acoustic emission event; defining an event epicenter as a single point where all the sagittal planes intersect with a plane perpendicular to the symmetry axis, and where said point represents a projection of the event hypocenter, and estimating the true hypocenter for an acoustic emission event from the epicenter by a 1D migration along a single line through the event epicenter along the symmetry axis.

Inventors:
HALDORSEN, Jakob B.U. (Elveveien 22, Nesbru, N-1394, NO)
Application Number:
EP2017/050446
Publication Date:
July 19, 2018
Filing Date:
January 11, 2017
Export Citation:
Click for automatic bibliography generation   Help
Assignee:
READ AS (Ravnsborgveien 56, Hvalstad, N-1395, NO)
International Classes:
G01V1/28
Foreign References:
US20120051178A12012-03-01
Other References:
VLADIMIR GRECHKA: "Tilted TI models in surface microseismic monitoring", GEOPHYSICS, vol. 80, no. 6, 22 May 2015 (2015-05-22), US, pages WC11 - WC23, XP055402951, ISSN: 0016-8033, DOI: 10.1190/geo2014-0523.1
ANDREW KING ET AL: "Anisotropy Effects on Microseismic Event Location", PURE AND APPLIED GEOPHYSICS ; PAGEOPH, BIRKHÄUSER-VERLAG, BA, vol. 164, no. 11, 3 November 2007 (2007-11-03), pages 2141 - 2156, XP019560424, ISSN: 1420-9136, DOI: 10.1007/S00024-007-0266-8
JAKOB B U HALDORSEN ET AL: "Locating microseismic sources using migration-based deconvolution", GEOPHYSICS, SOCIETY OF EXPLORATION GEOPHYSICISTS, US, vol. 78, no. 5, 28 August 2013 (2013-08-28), pages ks73 - ks84, XP001584714, ISSN: 0016-8033, [retrieved on 20130828], DOI: 10.1190/GEO2013-0086.1
ZHANG JIANZHONG ET AL: "Velocity modeling and inversion techniques for locating microseismic events in unconventional reservoirs", JOURNAL OF EARTH SCIENCE, CHINA UNIVERSITY OF GEOSCIENCES, HEIDELBERG, vol. 26, no. 4, 25 July 2015 (2015-07-25), pages 495 - 501, XP035519860, ISSN: 1674-487X, [retrieved on 20150725], DOI: 10.1007/S12583-015-0565-4
ALKHALIFAH, T.; LARNER, K., MIGRATION ERRORS IN TRANSVERSELY ISOTROPIC MEDIA: GEOPHYSICS, vol. 59, 1994, pages 1405 - 1418
ALKHALIFAH, T.; TSVANKIN, I.: "Velocity analysis for transversely isotropic media", GEOPHYSICS, vol. 60, 1995, pages 1550 - 1566
HALDORSEN, J.B. U.,; N.J. BROOKS; M. MILENKOVIC: "Locating microseismic sources using migration-based deconvolution", GEOPHYSICS, vol. 78, no. 5, 2013, pages KS73 - KS84, XP001584714, DOI: doi:10.1190/geo2013-0086.1
Attorney, Agent or Firm:
ONSAGERS AS (P.O. Box 1813 Vika, Oslo, N-0123, NO)
Download PDF:
Claims:
A method for fast determination of the location of a hypocenter of an acoustic emission event for a Vertical or Tilted Transverse Isotropic formation, each having a symmetry axis, by means of at least three receivers comprising the following steps: spreading the receivers out perpendicular to a symmetry axis for wave velocities; defining independent source/receiver sagittal planes as planes parallel to said symmetry axis that include the location of the receiver and

- either the particle movement polarization vector of compressional waves (P waves) or vertical shear waves (Sv waves) generated by the acoustic emission event,

- or the planes perpendicular to the particle movement polarization vector of a horizontal shear wave (Sh wave) generated by the acoustic emission event; defining an event epicenter as a single point where all the sagittal planes intersect with a plane perpendicular to the symmetry axis, and where said point represents a projection of the event hypocenter, and estimating the true hypocenter for an acoustic emission event from the epicenter by a ID migration along a single line through the event epicenter along the symmetry axis.

The method according to claim 1 , where in the case of a Tilted Transverse Isotropic formation, the sagittal plane in rotated to appear as a Vertical Transverse Isotropic, VTI sagittal plane prior to the estimating step.

The method according to claim 1 , where the single point representing a projection of the event hypocenter is defined from the intersection of straight lines in the horizontal plane directed along the observed polarization of compressional waves in the plane perpendicular to the symmetry axis.

The method according to claim 1 , where the single point representing a projection of the event hypocenter is defined from the intersection of straight lines along directions perpendicular to the polarization of shear waves in the plane perpendicular to the symmetry axis.

Description:
A method for fast determination of the location of an acoustic emission event in a VTI/TTI medium

Introduction The invention relates to the field of seismic prospecting, and more particularly concerns a method for fast determination of the hypocenter of an acoustic emission event for a vertical or tilted transverse isotropic (VTI/TTI) by means of micro- seismic monitoring.

Background

Micro-seismic monitoring may be used for estimating the size and orientation of induced fractures in formations. The main objective is to analyze data related to micro-seismic events. This includes their locations and the mechanisms associated with their generation. Micro-seismic activity can be detected by placing an array of geophones in a wellbore. By mapping the location of any small seismic events associated with a fracture, the approximate geometry of the fracture can be decided.

The accuracy of micro-seismic event mapping is dependent on the signal-to-noise ratio and the distribution of sensors. A Vertical Transverse Isotropic (VTI) formation is a formation where wave velocity only varies with depth, i.e. in the vertical direction. That is, the vertical axis is forming an axis of symmetry for the formation properties. This is for instance described in Alkhalifah, T., and Larner, K., 1994, Migration errors in transversely isotropic media: Geophysics, 59, 1405-1418 and Alkhalifah, T., and Tsvankin, I., 1995, Velocity analysis for transversely isotropic media: Geophysics, 60, 1550- 1566.

With the anisotropy of a VTI symmetric formation, where the velocities are invariant under lateral translation, ray bending will be entirely in a vertical plane, i.e. the sagittal plane of source/receiver. This also applies to a formation where the axis of symmetry is not vertical, i.e. the symmetry axis is tilted away from the vertical axis. This is called a Tilted Transverse Isotropic (TTI) formation.

In such formations, the ray bending is all in a plane along the tilted symmetry axis and consequently, all rays project to straight lines in a plane perpendicular to this tilted symmetry axis, i.e. similar to the horizontal plane for a VTI formation. The difference between a VTI and TTI formations is only in the orientation of the symmetry axis. The transform from one to the other can be achieved with a rotation of the coordinate system, i.e., orienting the axis of symmetry along the z (or

"vertical") axis.

The two elastic waves used for estimating the location, i.e. the hypocenter, of micro-seismic events are of two types: compressional waves (P waves), that have associated particle movement along the ray of propagation, and shear waves that have associated particle movement polarized in a direction perpendicular to the ray of propagation. For a VTI medium, one also normally separate the shear waves into two categories: horizontal shear waves (Sh waves) that have associated particle movement polarized in the horizontal plane and perpendicular to the ray of propagation, and vertical shear waves (Sv waves) that have associated particle movement in a vertical plane, perpendicular to the ray of propagation. Whereas, during ray bending in a VTI medium, the P and Sv waves will both maintain their polarization vector in the same vertical (sagittal) plane containing both the source and receiver locations, the Sh waves will be polarized perpendicular to this plane.

READ previously introduced a unique approach in handling micro-seismic data processing, ref. Haldorsen, J.B. U., N.J. Brooks, M. Milenkovic, 2013, Locating microseismic sources using migration-based deconvolution: Geophysics, 78, 5, pp. KS73-KS84. The former least-squares inversion approach that is over 100 years old has many flaws and limitations when applied to oil fields and geothermal micro- seismic data. READ takes a staged approach that treats full waveform micro- seismic data. A proprietary de-noising method can also be applied for increasing the signal-to-noise ratio.

Erroneous time picks lead to large errors in traditional event locations. READs new method eliminates the need for time picks and therefore reduces location

uncertainty.

If receivers are spread out in the direction perpendicular to the symmetry axis of the formation, the situation becomes more complicated. However, according to the invention, the event epicenter can nevertheless be found by a least-squares intersection of the planes parallel to the symmetry axis through the receiver location, containing either the particle movement polarization vector of the compressional or a Sv shear wave, or perpendicular to the particle movement polarization vector of a Sh shear wave generated by the micro-seismic event. The location of a seismic event away from the epicenter can be found by a ID migration along a single line through the event epicenter along the symmetry axis. Doing this, gives a particularly fast way of finding the location of an acoustic emission event from data recorded with an arbitrary spatial distribution of receivers. For one single receiver, one does not know whether the recorded wave field at the receiver represents a shear wave or a compressional wave. Therefore, for the process of finding the event epicenter, two possible sagittal planes are assigned to each receiver, one for each of the possible identifications of the elastic waves observed at each receiver as a shear or a compressional wave field. Any two planes have a single line of intersection. This means that using only two receivers will give two distinctly different event epicenters with equal probability, one for each of the two possible identifications of the elastic waves. Using three receivers, each of the two sets of possible sagittal planes will contain three planes, one through each receiver. Each pair of planes within each of the two sets of planes will intersect along lines parallel to the symmetry axis, each set thus defining three parallel lines. These three parallel lines will coincide, within the measurement uncertainty, for the correct choice of sagittal planes, and be statistically distinct for the erroneous choice. Thus, using at least three receivers, the correct identification of the wave modes as either shear or compressional can be made based on the closeness of the intersections of the associated sagittal planes.

Short description

The present invention is defined by a method for fast determination of the location of a hypocenter of an acoustic emission event for a Vertical or Tilted Transverse Isotropic formation, each having a symmetry axis, by means of at least three receivers comprising the following steps: spreading the receivers out perpendicular to a symmetry axis for wave velocities;

- defining independent source/receiver sagittal planes as planes parallel to said symmetry axis that include the location of the receiver and

- either the particle movement polarization vector of compressional waves (P waves) or vertical shear waves (Sv waves) generated by the acoustic emission event,

- or the planes perpendicular to the particle movement polarization vector of a horizontal shear wave (Sh wave) generated by the acoustic emission event;

defining an event epicenter as a single point where all the sagittal planes intersect with a plane perpendicular to the symmetry axis, and where said point represents a projection of the event hypocenter, and estimating the true hypocenter for an acoustic emission event from the epicenter by a ID migration along a single line through the event epicenter along the symmetry axis.

Further features of the method are defined in the claims.

Brief description of the drawings

The invention is further described in the detailed description below, in reference to the drawings by way of non-limiting example of embodiments of the subject disclosure, and where wherein:

Figure 1 shows multiple seismic events within the same time window;

Figure 2 shows an extracted event of the energy on each of the three components across an entire array consisting of 12 three-component receivers;

Figure 3 shows azimuth angles for each receiver;

Figure 4 shows the same event as in figure 3 after rotating the sagittal plane;

Figure 5a - 5c shows back-projection of the recorded data;

Figure 6 shows a plotting of the energy-weighted correlation function;

Figure 7 shows the epicenter defined by intersecting straight lines;

Figure 8 shows the sum of the squares of the distance from a point in the surface horizontal plane to all of the straight lines in the figure 7, and

Figure 9 shows a single trace giving a well-focused location directly below the event epicenter.

Detailed description The invention concerns a method for fast determination of the hypocenter for an acoustic-emission event for a VTI or a TTI formation.

In order to explain the present invention, a method for micro-seismic data processing by using a vertical array is first described according to READs previously introduced approach as mentioned above.

A wave mode separation technique to point the incoming compressional energy in the plane of the micro-seismic source is used. The second and third components search for the Sv and Sh (vertical and horizontal shear waves) and align themselves accordingly. Each X, Y, Z component is rotated to a new P, Sv, Sh system. No hodograms or time picks are used. Using a model for the propagation velocities for elastic waves in the formation, for each point in image space, we can reconstruct possible P and S time signatures for waves originating from this point. As a next step Haldorsen et al. 2013 apply a proprietary deconvolution operator,

de-convolving the reconstructed S-wave with the reconstructed -wave

(compressional wave) to create a shear wave origin time relative to P. For any point to be a likely origin point for the micro-seismic event, the de-convolved time series should have a maximum at the relative time equal to zero. The de-convolution algorithm used by Haldorsen et al. 2013, called semblance-weighted deconvolution or energy-weighted correlation, is designed to minimize noise and improve event location accuracy. The energy-weighted correlation is applied with a simultaneous back projection in order to map a recorded event - either to all possible locations in the 3D image space, or along a vertical plane already determined during the wave- mode separation process.

The method can be summarized as follows [Haldorsen et al., 2013] :

1. Identify and window events

2. Separate shear and compressional waveforms

3. Establish location map

a. align (shift by t P (x,z) - t P mm {x,z)) and stack compressional traces b. align (shift by t s (x,z) - t P mm {x,z)) and stack shear traces

c. apply deconvolution filter based on aligned and stacked

compressional traces applied to aligned and stacked shear traces d. map(x,z) = value at t=0 of deconvolved trace The key steps in the method and process of Haldorsen et al. (2013) is illustrated in the figures 1 to 6.

As mentioned in the background above, the difference between a VTI and TTI formations is only in the orientation of the symmetry axis. The transform from one to the other can be achieved with a rotation of the coordinate system, orienting the axis of symmetry along a vertical axis.

When going from a VTI symmetry to a TTI symmetry, the term horizontal plane should be replaced by a term describing a plane perpendicular to the axis of symmetry, and the term vertical plane should be replaced by term describing a plane along this axis. For convenience, for a TTI medium, we will choose to define the source/receiver sagittal plane as being along the axis of symmetry, and the event epicenter as the projection of the event location on a plane perpendicular to the axis of symmetry. The simplest way to implement this is by first rotating the coordinate axes such that the rotated vertical axis falls along the axis of symmetry. With this in mind, anywhere where we in the following mention VTI also applies to a TTI formation once the rotation of coordinate axes has been performed. Figure 1 shows multiple events within the same time window, ref. step 1 above. An event detector is used to extract events for the next step. The figure shows the vertical components (Z), followed by two sets of horizontal components (X, Y) recorded for a number of typical events.

Figure 2 shows an extracted event of the energy on each of the three components across an entire array of 12 three-component receivers, and where this is in coordinates pointing to the directions Up, East and North, ref. step 2 above.

Figure 3 shows the azimuth angles that each of the receivers needs to be rotated to have the second component pointing along the source/receiver sagittal plane. The minimum of the cost function (dark blue zones) for each receiver. For this vertical array, all receivers agree that the source is at an azimuth angle of 57 degrees clockwise from East. This direction defines a source/receiver sagittal plane common to all receivers, with the source located in this plane, ref. step 3 above.

Figure 4 shows the same event after rotation to have the P and S wave components in the sagittal plane through the source and the receiver. A third wave mode, probably a transverse shear is visible on the third component.

Figures 5a to 5c shows back-projection of the recorded data, based on polarization, the geometry and the velocities, for finding the P and S source signatures for events that possibly could have been generated at any point in the 3D image space. For this, one would preferably use the azimuth calculated above with 2D migration in the source/receiver sagittal plane. A full 3D migration approach to solve this estimation problem is also available.

Figure 5a shows the estimated P signature. Figure 5b shows the S signature that conceivably could have been generated at a specific point in this plane. To accept that there is an event at that point, they require the two signatures to be coherent (similar in waveform) and synchronous. A measure of these two conditions being satisfied is the proprietary energy- weighted correlation function shown in Figure 5c.

Figure 6 shows a plotting of the energy-weighted correlation function at time=0 for each of the cells in the source/receiver sagittal plane. A map of the hypocenter locations in this plane can be generated. The width of the maximum of the image gives a direct measure of location uncertainty.

One can make a map of the event locations and characterize their spatial

distribution, and the temporal development of spatially related structures or clusters of events.

If both P and S wave modes are observed at the various three-component receivers, the method described above will give accurate estimations of event hypocenters. If only one of the P and S wave modes are detected, the location of an event hypocenter would be essentially obtained from the curvature of the wave front alone, requiring a larger spread (a longer/wider array) for accuracy. If the receivers are spread out in the direction perpendicular to the symmetry axis of the formation, the situation becomes more complicated.

In a VTI formation, the ray bending is still entirely in the source-receiver sagittal plane, but the sagittal plane is different for each receiver. However, as the sagittal planes are purely vertical, they will each intersect with the horizontal plane along straight lines. These straight lines will intersect at a single point, which is the surface projection of the event location, i.e. the epicenter of the event.

For a VTI medium, this can be summarized as:

- The ray bending will be entirely in a vertical plane, i.e. the source/receiver sagittal plane.

- The projections on to a horizontal plane of all rays are straight lines.

- These straight lines intersect at the epicenter of the event.

Figure 7 illustrates this principle. The figure is based on real data recorded by three- component receivers distributed along the "double Λ" structure in the middle of the figure. The straight lines are lines along the horizontal vectors formed by the two horizontal components of the data. These straight lines represent the projections on to a horizontal plane of all rays. Based on these straight lines, the event epicenter can be estimated determined by their closest interception.

The present invention is defined by a method for fast determination of the location of the hypocenter for an acoustic emission event for a Vertical or Tilted Transverse Isotropic formation (VTI, TTI), each having a symmetry axis, by means of at least three receivers. The method comprises several steps.

The first step is spreading the receivers out perpendicular to a symmetry axis for wave velocities.

The next step is defining independent source/receiver sagittal planes as the planes to said symmetry axis that include the location of the receiver and either the particle movement polarization vector of compressional waves (P waves) or vertical shear waves (Sv waves) generated by the acoustic emission event,

or the planes perpendicular to the particle movement polarization vector of a horizontal shear wave (Sh wave) generated by the acoustic emission event.

With the receivers spread out perpendicular to the symmetry axis, the event epicenter can be found from the intersection of the straight lines along the compressional polarization in the plane perpendicular to the symmetry axis, or along directions perpendicular to the shear polarization in the plane perpendicular to the symmetry axis.

The next step is defining an event epicenter as a single point where all the sagittal planes intersect with a plane perpendicular to the symmetry axis, and where said point represents a projection of the event hypocenter.

According to one embodiment of the invention, where the formation is a TTI formation, the sagittal plane in rotated to appear as a VTI sagittal plane prior to the next estimating step.

According to one embodiment of the invention, the single point representing a projection of the event hypocenter is defined from the intersection of straight lines in the horizontal plane directed along the observed polarization of compressional waves in the plane perpendicular to the symmetry axis.

According to another embodiment of the invention, the single point representing a projection of the event hypocenter is defined from the intersection of straight lines along directions perpendicular to the polarization of shear waves in the plane perpendicular to the symmetry axis.

The location of the event away from the epicenter can be found by a ID migration along a single line through the event epicenter along the symmetry axis.

The last step of the inventive method is estimating the true hypocenter for an acoustic emission event from the epicenter by a ID migration along a single line through the event epicenter along the symmetry axis.

Doing this, gives a particularly fast way of finding the location of an acoustic emission event from data recorded with an arbitrary spatial distribution of receivers.

Figure 8 shows the sum of the squares of the distance from a point in the surface horizontal plane to all of the straight lines in the figure 7. This sum of squares has a well-defined minimum, which is the least-squares VTI solution to the event epicenter.

Having an estimate of the event epicenter, we can find the depth of the event below the event epicenter by applying, to all available data, a ID migration based on the principles described above, ref. [Haldorsen, et al, 2013].

Figure 9 shows an example of doing this to create a single trace giving a well- focused location directly below the event epicenter. The processing used is equivalent to the processing that gave the 2D image in the sagittal plane, described above with the processing used for a vertical array. One difference is that for a vertical array the migration needed to be 2D, whereas we now only need a ID image in the depth direction, as the event epicenter is already known. Another difference is that only P waves were used for this image, with the accuracy in the depth estimate essentially coming from the curvature for the P wave front, as measured over the aperture of the receiver array.

For a formation that is only approximately VTI/TTI, one can refine the estimates of event-hypo center locations by applying 3D migration to image the recorded data into a finely discretized 3D image using the method described above, ref.

[Haldorsen, et al., 2013].

With an arbitrary distribution of receivers, in any VTI/TTI symmetric formation, each receiver can be seen to be associated with its own source/receiver sagittal plane. According to the invention, this sagittal plane is defined as the plane that includes both the receiver and the source locations, and is parallel to the axis of symmetry. For a VTI symmetry, the sagittal planes are all vertical. For a TTI symmetry, the sagittal planes will appear as vertical after an appropriate TTI-to-VTI rotation. For any source/receiver pair, any ray bending between the source and the receiver will be entirely within its individual sagittal plane. All sagittal planes are independent. The key observation is that the intersections of all of these planes with a plane perpendicular to the symmetry axis, corresponding to the horizontal plane for a VTI system, will all meet in a single point. This point is the projection of the event hypocenter, i.e. the true 3D location, onto the plane perpendicular to the axis of symmetry. This point is defined as the event epicenter, and it can be estimated straightforwardly from the intersection of lines in the horizontal plane directed along the observed polarization of the compressional waves, which is oriented along the rays, or perpendicular to the observed polarization of the shear waves, which is oriented perpendicular to the ray.

Any axis of symmetry for anisotropy in geological formations are related to layering and stress applied to the formation. The main layering is horizontal and the main stresses in a geological formation is vertical, giving, predominantly, the vertical axis as the axis of symmetry. Higher-order symmetries, like orthorhombic or triclinic symmetries, may be introduced by vertical or sub-vertical fractures or horizontally applied stress, resulting in, e.g., shear velocities that vary with azimuth.

Within the framework and method according to the present invention, higher-order symmetries can be handled in secondary final full 3D estimates of location in a limited aperture, finely discretized 3D migration.