RU2784974C2 - Method and device for monitoring of subsoil under target area - Google Patents

Abstract

FIELD: seismic analysis.

SUBSTANCE: present invention relates in general to the field of seismic analysis and, in particular, to monitoring of subsoil and detection of anomalies in them, for example, detection of cavities under railways, automobile roads, buildings, as well as airports. For monitoring of terrestrial bowels under target area (3), seismic waves from identified mobile noise source (24) are recorded using at least one pair of sensors (22) located from both sides of target area (3), time periods are selected, corresponding to alignment of the pair of sensors (22) with noise source (24), using interferometry, a seismogram of target area (3) is restored based on recorded seismic waves and selected time periods, and an image of subsoil under target area (3) is generated based on the seismogram.

EFFECT: increase in the accuracy and information content of received data to generate an image of subsoil under a target area.

19 cl, 6 dwg

Description

The present application relates to a method and apparatus for monitoring the subsoil below the target zone.

The present invention generally relates to the field of seismic analysis and, in particular, to the observation of the subsoil and the detection of anomalies in the latter, for example, the detection of cavities under railways, roads, buildings, and airports.

To ensure the safety of surface infrastructure users, it is important to conduct instantaneous, temporary or permanent subsurface monitoring to detect any subsurface anomalies, such as cavities, that could weaken surface structures.

In an article titled "Seismic interferometry of railroad induced ground motions: body and surface wave imaging" published in Geophysical Journal International - 205, 2016, pp. 301-313, Diego A. QUIROS et al. propose a method for generating virtual propagation paths along a railroad track based on an interferometric analysis of ambient noise captured using a single receiver line or seismic wave sensors located along the railroad track.

The disadvantage of this method is the limited amount of information about the subsoil obtained directly under the line of sensors and does not allow not only to build a three-dimensional image of the subsoil under the railroad tracks, but also a two-dimensional image of the ground line parallel to the railroad track.

Also, in this purely passive method, train noise is mixed with other noise and applied when the latter are far enough away from the receivers to provide approximately the same noise level to the sensors on the line. Thus, the overall noise level is relatively low and does not allow imaging.

Papers "Near-surface velocity imaging using traffic induced high-frequency ground motion", H. NAKATA et al. (SEG 2017, p.2940-2945) and "Distributed acoustic sensing for seismic monitoring of the near surface: a traffic-noise interferometry case study" S. DOU et coll., Scientific Reports 2017; 7:11620 offer similar approaches with the same disadvantages of 2D imaging near areas to be explored.

The aim of the present invention is to eliminate the disadvantages of the prior art.

To this end, the present invention proposes a method for monitoring the subsurface under the target zone in accordance with claim 1.

In particular, seismic wave sensors are located on both sides of the target zone, and the identified source of seismic noise moves, if possible, outside the target zone, and even outside the area in which the sensors are installed. The method includes registration of waves coming from the source at certain periods by pairs of sensors, while the periods are selected so that the source and pairs of sensors are aligned, while the target zone is located between the pairs of sensors, then an interferometric reconstruction of the seismogram is performed based on the received data. Further, on the basis of the seismogram, an image of the subsoil under the target zone is built.

Thus, the use, on the one hand, of a mobile source of seismic noise, the characteristics of which are known due to the analysis of the signal itself and do not require a priori information, and on the other hand, at least two sensors that are aligned with this noise source for a certain period of time, allows carry out passive monitoring in space-time, namely, in four dimensions, with a large amplitude and high resolution, of that part of the subsurface that is located under the surface on which the sensors are placed, provided that the mobile noise source is aligned with a pair of sensors at least for a while.

Thus, a pair of sensors located on both sides of a target area (for example, a railway track) allows you to build an image of the subsurface under this target area, if a mobile sound source (for example, a train passing on a different track) moves close to the sensors and at least aligned with a pair of sensors for a while. Increasing the number of sensors around the target area, for example by placing them parallel to each side of a railroad track or even at the rails of a railroad track, allows a three-dimensional image of the target area to be generated due to the movement of the sound source.

In a specific embodiment, the method also includes the step of pre-processing the recorded seismic waves prior to the recovery step.

This allows you to receive signals of optimal quality, which increases the accuracy of the resulting image of the subsurface.

The advantage of the pre-processing step is that it includes noise removal and frequency filtering operations.

Additionally, the pre-processing step may include a spectral whitening operation and/or a binarization operation.

This type of operation makes it possible to simplify subsequent processing when reconstructing a seismogram.

In a specific embodiment, the interferometry reconstruction step performs at least one cross-correlation of the seismic waveforms recorded by the corresponding pair of sensors.

This allows you to effectively restore the seismogram.

In a particular implementation, the recording step is performed continuously in time.

This allows you to create reports on the status of the subsoil at a selected frequency, such as daily, weekly or monthly.

In a specific embodiment, the method, after the recovery step, further includes the step of summing the recovered seismogram signals from a plurality of mobile noise sources aligned with a respective pair of sensors at different time periods.

This summation step improves the quality and reliability of the recovery.

In a certain implementation, surface seismic waves are recorded during the recording step.

In this implementation, during the imaging phase:

surface wave dispersion curves are determined in the reconstructed seismogram signals in a given frequency range;

velocity tomography of the signals reconstructed from the dispersion curves and

the surface wave velocities of the reconstructed signals are inverted to obtain a P- and S-wave velocity model.

In a certain implementation, refracted seismic waves can be recorded during the recording step.

In this option, at the image generation stage:

the arrival time of the reconstructed seismogram signals corresponding to the refracted seismic waves is determined; and

a tomography of the reconstructed signals corresponding to refracted P-type waves is carried out based on the aforementioned signal arrival times so as to obtain a P-wave velocity model.

Depending on the specific characteristics, it is possible to register reflected seismic waves, the sensors may include at least one geophone and / or at least one accelerometer and / or at least one sensor based on the use of fiber optic cables (for example, designed for distributed acoustic sensing ), the sensors should be located at least in two lines, or in the nodes of a grid with square or hexagonal cells and/or may be located at equal distances.

The aforementioned target area may be a railroad track, road or other engineering structure or building.

The target area may be one railway track, and the noise source may be a train moving on a second track that runs parallel to the first. The sensors may be located at regular intervals on two imaginary lines located on both sides of the first railway track and running parallel to it.

With the same purpose as above, the present invention also provides a device for monitoring the subsurface below the target zone. Its distinguishing feature is that it includes the elements necessary to carry out the steps in the process outlined above.

Since the individual characteristics and advantages of the apparatus are similar to those of the method, they are not repeated here.

Other aspects and advantages of the invention will become clear upon reading the following detailed description of specific embodiments, made in the form of non-limiting examples with reference to the accompanying figures, in which:

- figure 1 shows a block diagram with the stages of the process of monitoring the subsurface under the target zone in accordance with the present invention in a particular embodiment;

- figure 2 is a schematic view illustrating the configuration for recording seismic waves in a non-limiting example, where the target area is a railway track;

- Figure 3 is a diagram illustrating a configuration for restoring virtual distribution paths for the example of FIG. 2;

- Figure 4 is a diagram illustrating a configuration for restoring virtual multi-azimuth propagation paths for the example of FIG. 2;

- Figure 5 shows a configuration diagram for restoring virtual propagation paths over multiple azimuths and using multiple railroad tracks; and

- figure 6 shows a diagram of a preferred embodiment of the invention.

As shown in the flow chart in figure 1, during the first stage 10 of the process of monitoring the subsurface under the target zone according to the invention, at least for a certain period of time, seismic waves are recorded.

These seismic waves come from a moving source of seismic noise, that is, from a source that is in motion and whose movement creates noise. This noise source is known in the sense that it is well identified. Thus, we are not talking about ambient noise of unidentified origin.

Seismic wave data is recorded using a pair of sensors positioned to be aligned with the noise source for a specified period of time.

Pairs of sensors can be placed in a variety of ways, such as at regular intervals in straight lines, as shown in Figure 2 below, but they can also be placed irregularly and along any curve, or at regular or irregular grid points with square cells. hexagonal or otherwise. They can also be placed randomly.

Figure 2 shows, as one non-limiting example, a sensor configuration for recording if the target area is a railroad track 20. However, the invention is also well applicable to a target area that includes a conventional road, building, or other infrastructure (for example, an airport) .

A plurality of seismic wave receivers or sensors 22 are located along the track 20, referred to as the "target", on either side of it. These sensors 22 form "Receive Line A" 22A and "Receive Line B" 22V. DAS (Distributed Acoustic Sensing System).

Along the target path 20, the sensors 22 may be distributed at regular intervals (eg, several meters apart). In particular, the distance between the sensors 22 may be 2 meters.

The sensors 22 may be placed on the surface of the earth or slightly buried, namely several meters below the surface. They can be placed on/under a railway embankment, on or under rails, and in a railway tunnel.

The plurality of sensors 22 includes at least one pair of sensors 220, 221 located opposite each other on both sides of the railroad track 20 and transverse to it.

The mobile source of seismic noise is a train 24 moving along track 21 - the so-called "source track" - separate from the target track 20. The noise emitted by the moving train 24 is symbolized in the figure by arrows in the form of circles 26. This noise mainly arises from contact of the wheels of the train 24 with the rails of the railway track 21.

A set consisting of a plurality of sensors 22, a mobile seismic noise source 24, and a module suitable for carrying out processing of the subsurface monitoring method described herein forms a subsurface monitoring device according to the invention.

Considering a pair of sensors 220, 221, they will be in alignment with train 24 for a period of time that starts when the first wheel of train 24 is on a straight line passing through the barycenter of sensors 220 and 221 and ends when the last wheel train 24 is on a straight line passing through the barycenter of sensors 220 and 221.

In accordance with the present invention, during this period of time, the sensors 220 and 221 register seismic waves, which will be subjected to further processing by the method corresponding to the invention. This period of time is predetermined because the kinematic parameters of the train 24 are known, namely its position and speed at each moment. In this sense, the mobile source of seismic noise, which is the train 24, is identified.

The recording step 10 may be limited to a given time period. In one embodiment, registration may also be continuous in time.

As shown in the flowchart in Figure 1, the seismic acquisition step 10 may be followed by a pre-processing step 11 for the seismic waves detected by the sensors 22.

This pre-processing can be carried out in various ways and includes at least a noise removal step and a frequency filtering step.

In one of the described examples, the denoising step of the recorded seismic waves removes the baseline ambient noise from the seismic noise produced by the train 24. This baseline ambient noise is electrical noise associated with the catenary network, the characteristics of which are well known. The result is a signal with an improved signal-to-noise ratio.

During the frequency filtering stage, the signal is filtered in a certain frequency range that corresponds to the expected emission band of the train 24 and the target emission band, which depends on the depth of the subsurface sounding under the target area represented by the railway track 20.

In addition, in between the removal step and the frequency filtering step, a spectral whitening operation and/or a binarization operation may additionally be performed.

The operation of spectral whitening, also known as spectral equalization, consists in weighting all the frequency components of a signal so that they all have the same energy representation.

The binarization operation, also commonly known, is to assign -1 or +1 to each instantaneous signal value in order to simplify the signal.

Other types of pretreatment may also be used.

Step 10 (registration) and optional step 11 (pre-processing of seismic waves registered by sensors 22) are followed by step 12 - recovery of the seismogram of the target zone.

In recovery step 12, the interferometry method is applied to the recorded and possibly pre-processed seismic waves.

This interferometry method includes at least one operation of cross-correlation of signals obtained from seismic waves recorded by sensors 220 and 221 during a certain period of time during which the train 24 passed along the source track 21. This interferometry method allows you to restore the virtual propagation paths of transverse waves, in this case perpendicular to the target railroad track 20. These virtual propagation paths characterize the seismogram.

Figure 3 schematically shows the virtual propagation paths for each pair of sensors located on either side of the target track 20.

In other embodiments, any other interferometry method, such as a convolution-based method, may be applied in recovery step 12 instead of the cross-correlation operation mentioned above.

As shown in figure 4, the restoration step 12 can be performed in different directions. Indeed, cross-correlation operations can be performed not only between seismic waveforms recorded by pairs of sensors located opposite each other (like a pair of sensors 220 and 221), but also pairs of sensors located similarly to sensors A5 and B6 or sensors A7 and B6 , that is, pairs of sensors for which the straight line passing through the barycenter of the two sensors included in the pair forms a non-zero angle with the perpendicular to the target railway track 20.

Moreover, as shown in figure 5, the recovery step 12 can be carried out not only for different directions, but also using a plurality of railway tracks with corresponding pairs of sensors. Source path 21, parallel to paths 201-204, runs adjacent to "receiver line 22A" as in Figures 2-4, but is not shown in Figure 5 for simplicity.

Thus, the passage of the train along the source track 21 makes it possible to reconstruct, using interferometry, the virtual propagation paths between lines 22A and 22B, and/or between lines 22B and 22C, and/or between lines 22C and 22D, and/or between lines 22D and 22E. In other words, for each pair of adjacent railroad tracks, each track allows the seismogram of the zone located under the adjacent track to be reconstructed using pairs of sensors deployed on either side of this adjacent track.

The procedure for restoring virtual transverse propagation paths (i.e., transverse to the railroad track) has been described above using a time window during which train 24 is aligned with a pair of sensors 220, 221: the pairs of sensors allow the study of seismic noise path variation.

In one embodiment, it is possible to use convolution correlation between two sensors on either side of a track and a signal emitted by a train traveling on the same track. Thus, temporal variations in the propagation time can be used for continuous monitoring in order to track the occurrence of anomalies.

As shown in Figure 1, the recovery step 12 may optionally be followed by a summation step 13 of the recovered seismic waveforms received from a plurality of mobile seismic noise sources 24.

In the non-limiting example described here, the noise source is trains, a plurality of trains traveling along a track at different times, having different lengths and speeds, and accordingly aligned with sensor pairs like pair 220, 221 at different times. Such trains will give similar restored signals.

The summation of several such signals allows to improve the signal-to-noise ratio and, thus, to obtain a better image of the subsoil under the target zone. It also allows sensors such as the pair 220, 221 to recreate the virtual propagation paths once or at predetermined intervals such as daily, weekly or monthly.

As shown in FIG. 1, after the recovery step 12 and the optional summation step 13, a step 14 of subsurface imaging under the target zone is performed based on the seismogram recovered earlier.

In imaging step 14, the processing depends on the type of seismic waves recorded in step 10. In fact, different types of waves can be recorded: surface waves, refracted waves, and also reflected waves.

In one embodiment, in which surface seismic waves are recorded in step 10, the following operations are performed to obtain an image of the subsurface under the target zone.

First of all, the dispersion curves of surface waves in the reconstructed seismogram signals in a given frequency range are determined. As a non-limiting example, this frequency range may be in the range from 1 to 100 Hz.

If necessary, you can then select only the maxima of these dispersion curves.

Then, velocity tomography of the reconstructed signals is performed based on the dispersion curves.

Finally, the surface wave velocities of the reconstructed surface wave signals are inverted to obtain a model of the pressure and shear wave velocities, also known as P- and S-waves.

By repeating these operations over time, an image of the subsurface under the target zone is obtained that changes over time, in other words, a four-dimensional representation of the given subsurface area is obtained.

In another implementation, in which refracted seismic waves are recorded in step 10, the following operations are performed to obtain an image of the subsurface under the target zone.

First of all, the arrival time of the reconstructed seismogram signals corresponding to refracted seismic waves is determined.

Next, the reconstructed signals corresponding to refracted P-waves are tomographed based on the aforementioned signal arrival times in order to obtain a velocity model of pressure waves, also known as P-waves.

Similar to the seismic surface wave implementation, repeating these operations over time produces a time-varying image of the subsurface below the target zone—in other words, a four-dimensional image of that subsurface area.

In another embodiment of the invention, at step 10 it is possible to register reflected seismic waves. It allows you to get a picture of the reflectivity of the subsurface under the target zone.

Of course, it is possible to combine the registration of refracted waves with the registration of surface waves in order to increase the resolution, as well as to integrate the processing of reflected waves into these processes. In particular, the signals are processed independently to peaks that are then combined, with the dispersion curves for surface waves and the first arrival times for refracted waves acting as peaks.

Particularly preferably, and as shown in FIG. 6, an image of an underground target area 3 under the railway track 20, about 1.5 m wide, lying some distance from the parallel source track 21, can be obtained using two lines of conventional sensors such as accelerometers, spaced 5 m apart and spanning the track 20, the distance between the sensors in each line is 3 m, the sensors are staggered to equalize the resolution of the three-dimensional image obtained by cross-correlation on the gauge scale. Processing of surface and refracted waves can achieve a resolution of 2 m or less. If a higher resolution is required, it is possible to bring the sensor lines closer together, for example, by placing them right next to the rails, i.e. at a distance of about 1.5 m, and/or use two fiber optic cables for distributed acoustic sensing, thereby reducing the distance between the sensors to the order of 60 cm

To the method described in the present invention, it is also possible to add a prior art reconstruction method in which the sensor arrays make it possible to analyze the seismic wave dispersion curves to further improve the resolution and/or fidelity of the image: namely, the virtual longitudinal propagation paths are restored, i.e. parallel to the railway track, for example, using time windows of predetermined duration before and after the passage of train 24, respectively, after the sensors begin to register seismic waves and before the sensors stop registering seismic waves.

The method according to the invention involves passive monitoring of the subsurface below the target zone in the sense that it processes the received seismic waves but does not produce them. However, it is possible to imagine combining seismic sensors with active seismic sources to apply the processing methods described in US-A-2017 052269.

Claims (28)

1. A method for monitoring the subsoil under the target zone, including the following steps, in which: registering (10) seismic waves from an identified and moving source (24) of seismic noise using at least one pair (220, 221) of sensors (22) located on both sides of the specified target zone; select time periods corresponding to the alignment of the pair (220, 221) of the sensors (22) and the specified noise source (24); restore (12) using interferometry seismogram of the target zone based on the specified registered seismic waves and selected time periods and generating (14) an image of the subsurface under the target zone based on the specified seismogram. 2. The method according to claim 1, characterized in that the source (24) is moved outside the segment formed by the location of said pair of sensors (220, 221). 3. The method according to claim 1 or 2, characterized in that it further comprises, prior to the recovery step (12), a step (11) of pre-processing the recorded seismic waves. 4. Method according to claim 3, characterized in that the pre-processing step (11) includes noise removal and frequency filtering operations. 5. The method according to claim 3 or 4, characterized in that the pretreatment step (11) includes a spectral bleaching operation. 6. The method according to any one of the preceding claims, characterized in that during the step (12) of restoration using interferometry, at least one operation of cross-correlation of signals received from seismic waves, respectively recorded by two sensors from the specified pair of sensors (220, 221 ). 7. Method according to any one of the preceding claims, characterized in that step (10) of registration is carried out continuously in time. 8. The method according to any one of the preceding claims, characterized in that it further comprises, after the step (12) of recovery, the step (13) of summing the recovered signals of the specified seismogram from a plurality of mobile noise sources, respectively aligned with the aforementioned pair (220, 221) of sensors during different periods of time. 9. A method according to any one of the preceding claims, characterized in that at the recording step (10) surface seismic waves are recorded. 10. The method according to claim 9, characterized in that at the stage (14) of generating the specified image: determine the dispersion curves of surface waves in the recreated signals of the specified seismogram in a given frequency range; conducting tomography of the speed of the above signals, restored on the basis of the above dispersion curves; and inverting the surface wave velocities of the aforementioned recovered signals in order to obtain a P- and S-wave velocity model. 11. The method according to any one of paragraphs. 1-10, characterized in that at the registration stage (10) refracted seismic waves are recorded. 12. The method according to claim 11, characterized in that in step (14) of generating the above image: determining the arrival time of the recovered signals of said seismogram corresponding to said refracted seismic waves; and performing tomography of said reconstructed signals corresponding to refracted P-type waves based on said signal arrival times in order to obtain a P-wave velocity model. 13. A method according to any one of the preceding claims, characterized in that at the recording step (10) reflected seismic waves are recorded. 14. A method according to any one of the preceding claims, characterized in that said sensors (22) comprise at least one geophone and/or at least one accelerometer and/or at least one sensor based on the use of optical fibers. 15. A method according to any of the preceding claims, characterized in that said sensors (22) located on both sides of said target area are aligned in at least two lines, or placed at the nodes of a grid with square or hexagonal cells, and/or spread evenly. 16. Method according to any one of the preceding claims, characterized in that said target zone is a railway (20), a road or a building. 17. Method according to any one of the preceding claims, characterized in that said target zone (3) is a first rail track and said source (24) is a train moving on a second rail track (21) substantially parallel to the first rail track (twenty). 18. The method according to claim 17, characterized in that said sensors (22) are evenly placed on two parallel lines surrounding the first railway track (20). 19. Subsoil monitoring device under the target zone, characterized in that it contains at least one pair (220, 221) of sensors (22) located on both sides of the specified target zone, a movable seismic noise source (24) and a module made with the possibility of performing processing operations of the method according to any of the previous paragraphs.

Images