Pit and well abnormal location detection device and abnormal location detection method

Abstract

To provide a detection device and a detection method that detects abnormal flows such as the outflow of mine water and detects abnormal points such as cracks in a well wall by using observation images from inside a geothermal well.SOLUTION: Video data is captured by a borehole camera, the data is divided into frames, suspended particles in each frame are detected, the movement of a specific suspended particle in each frame is detected, a moving velocity vector V(θ,t) is calculated, and the moving velocity |V(θ,t)| is calculated from the moving direction θ(t) at time t and V(θ,t). A descent velocity |Vz(t)| is calculated based on a depth Dz(t) of a sonde part of the borehole camera, V(θ,t) is divided by |Vz(t)| to calculate a normalized velocity vector P(θ,t) of the suspended particles, a weighting coefficient f is calculated from P(θ,t), the position of the suspended particles and P(θ,t) are multiplied by f, a weight-corrected normalized velocity spectrum P'(θ,t) for each time t is calculated, and the presence or absence of an abnormality in the well is determined based on P'(θ,t) at time t.SELECTED DRAWING: Figure 1

Description

Technical Field

[0001] The present invention relates to an abnormal location detection device and an abnormal location detection method for a mine shaft that can detect abnormal flows such as the outflow of underground water and the inflow of hot water from the observation video in the mine shaft, and can detect abnormal locations such as cracks in the mine wall.

Background Art

[0002] Conventionally, in order to utilize geothermal energy and hot springs as energy and tourism resources, a heat exchanger is inserted into an excavation pit (geothermal well or hot spring well) dug on the ground surface to extract heat from the ground, or hot water (including hot spring water) or steam is directly collected from underground through an excavation pit (production well), and the water after being used for power generation on the ground is reduced into the ground through an excavation pit (geothermal reinjection well). Also, water is injected into the geothermal well to collect the water heated by geothermal energy or the generated steam.

[0003] In hot spring wells, the depth may exceed 1000 m, and in geothermal wells, it may exceed 2000 m. In particular, the temperature inside geothermal wells is often in a harsh environment exceeding 200°C. Also, the diameter of geothermal wells and hot spring wells is only about several tens of cm even at the wellhead, making it extremely difficult to inspect inside geothermal wells and hot spring wells.

[0004] Therefore, by inserting a camera called a borehole camera as disclosed in Patent Document 1 into a geothermal well and photographing the inner surface of the geothermal well, observations are made on cracks in the well wall, the state of faults, and protective pipes (casing) for protecting the well.

Prior Art Documents

Patent Documents

[0005]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0006] However, the inside of a geothermal well or a hot spring well stores well water, and this well water contains suspended particles such as, for example, fragments of rock mass, peeling scale from the well wall, minerals contained in hot water, and microorganisms. When such suspended particles are present, the video taken by a borehole camera becomes blurred, or it becomes difficult to observe the state of the inner surface of the geothermal well or the hot spring well.

[0007] Therefore, whether a human visually observes the video taken by the borehole camera or computer analysis is performed, the suspended particles become an obstacle to observation, and there is a possibility of overlooking cracks in the well wall or damaged portions of the protection pipe of the well.

[0008] In view of such a current situation, an object of the present invention is to provide an abnormal location detection device and an abnormal location detection method for a well that can detect abnormal flows such as the outflow of well water and the inflow of hot water from the observation video in the well and detect abnormal locations such as cracks in the well wall.

Means for Solving the Problem

[0009] The present invention was invented to solve the problems in the prior art as described above, and the abnormal location detection device and the abnormal location detection method for a well of the present invention include those configured as follows.

[0010] [1] An abnormal location detection device for detecting an abnormal location of a well, detecting a predetermined number of suspended particles in a predetermined frame from video data taken by a borehole camera connected to the abnormal location detection device and Of the said shaft calculating the center position, Regarding the detected predetermined number of suspended particles as specific suspended particles by detecting the Specify suspended particles in the frame next to the said frame, Specify detecting the movement of the suspended particles and calculating a movement speed vector V(θ,t), the SpecifyFor the floating particles, the moving speed |V(θ,t)| is calculated from the moving direction θ(t) and the moving speed vector V(θ,t) at time t, and Based on the depth Dz(t) of the sonde part of the borehole camera, the descending speed |Vz(t)| of the sonde part at time t is calculated, the Specify By scalar dividing the moving speed vector V(θ,t) of the floating particles by the descending speed |Vz(t)| of the sonde part, the Specify normalized speed vector P(θ,t) of the floating particles is calculated, The center position of the said shaft, the Specify position of the floating particles And normalized speed vector P(θ,t) And to calculate the weighting coefficient f, the Specify By multiplying the normalized speed vector P(θ,t) of the floating particles by the weighting coefficient f, the weighted corrected normalized speed spectrum P’(θ,t) for each time t is calculated, An abnormal location detection device configured to determine whether there is an abnormality in the shaft based on the weighted corrected normalized speed spectrum P’(θ,t) at time t.

[0011] [2] An abnormal location detection method for detecting an abnormal location in a shaft, comprising: Shooting video data with a borehole camera, Detecting a predetermined number of floating particles in a predetermined frame from the video data and Of the said shaft calculating the center position, Regarding the detected predetermined number of suspended particles as specific suspended particles In the next frame of the frame, the Specify floating particles are detected, Specify to detect the movement of the floating particles and calculate the movement speed vector V(θ,t), the Specify For the floating particles, the moving speed |V(θ,t)| is calculated from the moving direction θ(t) and the moving speed vector V(θ,t) at time t, and Based on the depth Dz(t) of the sonde part of the borehole camera, the descending speed |Vz(t)| of the sonde part at time t is calculated, The Specify normalized velocity vector P(θ,t) of the floating particles is calculated by scalar division of the movement velocity vector V(θ,t) of the floating particles by the descent velocity |Vz(t)| of the sonde section, Specify and The center position of the said shaft, the Specify position of the floating particles And and the normalized velocity vector P(θ,t) And are used to calculate the weighting coefficient f. The Specify weighted corrected normalized velocity spectrum P’(θ,t) for each time t is calculated by multiplying the normalized velocity vector P(θ,t) of the floating particles by the weighting coefficient f, and an abnormal location detection method for determining whether there is an abnormality in the shaft based on the weighted corrected normalized velocity spectrum P’(θ,t) at time t.

Advantages of the Invention

[0012] According to the present invention, by detecting abnormal flows such as the outflow of in-pit water and the inflow of hot water from the movement direction of floating particles floating in the in-pit water, abnormal locations of the shaft such as cracks in the shaft wall and damage to the protective pipe of the shaft, and feed points (water path locations with the geothermal reservoir) can be detected.

Brief Description of the Drawings

[0013]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

Figure 10

Figure 11

Figure 12

DETAILED DESCRIPTION OF THE INVENTION

[0014] Hereinafter, embodiments (examples) of the present invention will be described in more detail with reference to the drawings. FIG. 1 is a schematic diagram for explaining the configuration of the abnormal location detection device for a shaft in the present embodiment.

[0015] As shown in FIG. 1, the abnormal location detection device 10 for a shaft in the present embodiment includes a sonde section 20 inserted into the shaft 100, an armored cable 30 for suspending the sonde section 20 and communicating the information sent from the sonde section 20, and a winch 42 for unwinding and winding the armored cable 30 to lower and raise the sonde section 20, and is connected to a borehole camera 12 via the armored cable 30.

[0016] In the present embodiment, a camera 21 is provided vertically below the sonde section 20 and is configured to photograph vertically below the sonde section 20.

[0017] Such a borehole camera 12 is not particularly limited as long as it can photograph the video inside the shaft 100. For example, a commercially available borehole camera such as the high-temperature borehole camera system WTC-1500HT manufactured by Murata Manufacturing Co., Ltd. can be used.

[0018] Note that in this embodiment, the video captured by the borehole camera 12 is a video vertically downward with respect to the sonde unit 20. However, as will be described later, it may be a video in the horizontal direction with respect to the sonde unit 20. In this case, the camera 21 may be provided in the circumferential direction of the sonde unit 20, or by using the camera 21 provided vertically downward of the sonde unit 20 shown in FIG. 1 as an ultra-wide-angle camera, it can also be configured to be able to capture a video in the horizontal direction with respect to the sonde unit 20. Note that when using such an ultra-wide-angle camera, since distortion occurs in the captured video, it is preferable to perform distortion correction.

[0019] Further, when lowering and raising the sonde unit 20, it is preferable to provide a centering device 22 in the sonde unit 20 so that the sonde unit 20 is located at approximately the center of the shaft 100. By providing the centering device 22 in this way, when capturing a video vertically downward with respect to the sonde unit 20 as in this embodiment, the center of the captured video data and the center of the shaft 100 will approximately coincide.

[0020] The abnormal location detection device 10 can receive the video data sent from the borehole camera 12 and can be configured by a computer capable of analyzing the received video data as will be described later.

[0021] The abnormal location detection device 10 of this embodiment configured as described above can detect the abnormal location of the shaft as follows. FIG. 2 is a flowchart for explaining the process of detecting the abnormal location of the shaft 100 by the abnormal location detection device 10 of this embodiment.

[0022] First, the sonde unit 20 is inserted into the shaft 100, and the winch 42 is controlled to unwind the armored cable 30 to lower the sonde unit 20 into the shaft 100 (S10).

[0023] In this embodiment, during the descent of the sonde unit 20, the area below the sonde unit 20 is photographed (S20). Note that the timing of photographing the area below the sonde unit 20 is not limited to during the descent of the sonde unit 20, and for example, it may be performed at the timing when the sonde unit 20 stops.

[0024] The photographed video data is transmitted from the sonde unit 20 to the abnormal location detection device 10, and the abnormal location detection device 10 analyzes the received video data as follows. The abnormal location detection device 10 detects a predetermined number of floating particles in each frame (still image for each time) of the video data, and also performs tracking of specific floating particles (detection of the movement of specific floating particles in each frame) (S30).

[0025] Here, as a method for obtaining the flow velocity distribution in a liquid, the Particle Image Velocity (PIV) method is known. In the PIV method, particles (tracer particles) serving as indicators are mixed into the fluid, and by irradiating with strong light such as a laser, the tracer particles are visualized, and the flow velocity distribution is obtained from the movement amount of the particles between two micro times.

[0026] In the present invention, floating particles in a mine shaft that originally block vision are used as tracer particles to grasp the flow situation of the mine shaft water. Note that as a method for estimating the movement destination of the floating particles, for example, the position change between floating particles detected using background removal in different frames may be calculated, or an algorithm such as YOLO (You Only Look Once) may be used.

[0027] When detecting floating particles using the background removal method, as a pre-step for performing the detection process, for example, filter processing for reducing the contrast of an image such as compressing the histogram of luminance values may be performed, or a statistical distribution of the sizes of the detected floating particles may be examined, and processes such as excluding floating particles with a predetermined value or more may be performed to prevent detecting objects other than floating particles.

[0028] In addition, since the captured video data contains a large number of floating particles, for example, when detecting and tracking only a small number of floating particles in the video data, malfunction or detection omission is likely to occur. Therefore, in the present embodiment, in order to detect and track as many floating particles as possible included in the captured video data, not only between adjacent frames but also the movement speed vectors of the floating particles detected within a certain period of time are utilized. By detecting and tracking a large number of floating particles in this way, a large amount of data on flow vectors as described later can be acquired, the influence of malfunction and detection omission can be reduced, and the influence on the determination of abnormality can be reduced.

[0029] Next, the abnormality location detection device 10 calculates the moving direction and moving speed in each frame for a specific floating particle being tracked (S40). As shown in FIG. 3, when a specific floating particle existing at position P1 at time t moves to position P2 at time t+T int and the moving distance is d(t), the moving speed |v(t)| = d(t) / T int , and the moving direction θ(t) can be represented. Here, T int indicates the time difference between each frame. Also, the flow vector of a specific floating particle (the vector of the floating particle moving in the moving direction θ(t) and moving speed |v(t)|) is defined as the moving speed vector V(θ,t).

[0030] Note that depending on the video data, generally, there are dozens of frames between time t and time t+T int . From such a plurality of frames, two frames may be selected to calculate the moving speed vector V(θ,t), or a large number of moving speed vectors V(θ,t) that can be obtained from dozens of frames may be calculated and utilized. By using such a large number of moving speed vectors V(θ,t), the influence of detection omission and the like can be reduced, and abnormality detection can be performed accurately and easily.

[0031] Further, the abnormality location detection device 10 obtains the descending speed |Vz(t)| of the sonde unit 20 at time t based on the depth Dz(t) of the sonde unit 20 (S50).

[0032] Then, the abnormality location detection device 10 corrects the movement speed vector V(θ,t) of specific suspended particles by scalar division with the descending speed |Vz(t)| of the sonde unit 20, and calculates the normalized speed vector P(θ,t) of the specific suspended particles (S60). Since the flow of the suspended particles is affected by the descent of the sonde unit 20, in this way, by scalar dividing the movement speed vector V(θ,t) of the suspended particles by the sonde descent speed |Vz(t)|, the influence of the sonde descent speed can be corrected.

[0033] Also, the abnormality location detection device 10 calculates a coefficient f for weighting from the normalized speed vector P(θ,t) of specific suspended particles (S70). Specifically, when a specific suspended particle existing in the direction of angle φ from the center of the frame moves in the direction of angle θ, as shown in Fig. 4(a), when 0≦φ<+π and 0≦θ<+π, or as shown in Fig. 4(b), when -π≦φ<0 and -π≦θ<0, the coefficient f = cos(θ - φ); as shown in Fig. 5(a), when 0≦φ<+π and -π≦θ<0, or as shown in Fig. 5(b), when -π≦φ<0 and 0≦θ<+π, the coefficient f = cos{π - (θ - φ)} can be set. Here, when the coefficient f>0, the specific suspended particle is moving outside the shaft, and there is a high possibility that the water in the shaft is flowing out. On the other hand, when the coefficient f<0, the specific suspended particle is moving towards the center of the shaft, and there is a high possibility that hot water is flowing into the shaft.

[0034] Next, by multiplying the normalized speed vector P(θ,t) by the weighting coefficient f, for the "suspended particle in the direction of angle φ from the center of the shaft", the movement direction φ and the normalized speed vector P(θ,t)×f = P’(θ,t) are calculated, and the weighted corrected normalized speed spectrum P’(θ,t) for each time t is obtained. (S80)

[0035] In this embodiment, the direction is defined such that the upper side of the video data is 0, and it is shown as ranging from 0 to -π counterclockwise and from 0 to +π clockwise.

[0036] When there is a location where the spectral intensity (amplitude) exceeds a predetermined threshold value in the weight-corrected normalized velocity spectrum P’(θ,t), the abnormality detection device 10 determines it as an abnormality. (S90)

[0037] Specifically, when the weight-corrected normalized velocity spectrum P’(θ,t1) at time t1 is represented as shown in Fig. 6(a), if it falls within the threshold range as shown in Fig. 6(a), it is determined that there is no movement of suspended particles due to the inflow and outflow of in-pit water. In this case, due to the influence of the descent of the sonde unit 20, as shown in Fig. 6(b), it means that the suspended particles are spreading and moving outward from the center of the shaft 100.

[0038] When shooting video data with the sonde unit 20 stopped, after the sonde unit 20 is stopped, the state where the in-pit water becomes turbid due to contact with the shaft wall during the descent of the sonde unit 20 or the suspended particles that are wound up becomes dominant, and it is impossible to accurately grasp the flow of the in-pit water. After a while, when the winding up of these suspended particles subsides, the movement of the suspended particles comes to dominate the in-pit water flow. At that time, generally, hot water at a high temperature gushes up in the shaft, so compared with the pattern during descent, the state is at a lower level (the average value μ of the weight-corrected normalized velocity spectrum P’(θ,t) described later decreases), but basically a pattern similar to Fig. 6(b) is observed. Therefore, similar to the case of shooting video data while lowering the sonde unit 20, an abnormality can be judged.

[0039] Also, when the weight-corrected normalized velocity spectrum P’(θ,t2) at time t2 exceeds the threshold range as shown in Fig. 7(a), it is determined that the suspended particles are moving due to the inflow and outflow of the in-pit water. In this case, as shown in Fig. 7(b), it means that the suspended particles are moving in a specific direction toward the outside of the shaft 100, that is, the in-pit water is flowing out.

[0040] On the other hand, even when the weighted corrected normalized velocity spectrum P’(θ, t3) at time t3 is below the threshold range as shown in Fig. 8(a), it is determined that floating particles are moving due to the inflow and outflow of in-pit water. In this case, as shown in Fig. 8(b), the floating particles are moving in a specific direction toward the inside of the shaft 100, which means that hot water is flowing in.

[0041] When the sonde unit 20 is stopped to capture video data, it is necessary to wait until the swirling of the floating particles as described above subsides. However, when there is an outflow of in-pit water or an inflow of hot water, it will show a pattern similar to Figs. 7 and 8. Therefore, similar to the case of capturing video data while lowering the sonde unit 20, an abnormality determination can be made.

[0042] The threshold value is not particularly limited and can be set as appropriate. For example, when the average value of the weighted corrected normalized velocity spectrum P’(θ, t) for each time t is μ, the threshold value can be set to μ ± 2σ. Here, σ represents the standard deviation.

[0043] When the abnormality detection device 10 determines that an abnormality exists, it calculates the depths Dz(t2) and Dz(t3) of the sonde unit 20 at times t2 and t3, and determines that an abnormality has occurred in the shaft at the depths Dz(t2) and Dz(t3) (S100).

[0044] The analysis result calculated by the abnormality detection device 10 is preferably configured to be notified to the user, for example, by displaying it on a display device (not shown) of the abnormality detection device 10.

[0045] Also, in the above embodiment, detection and tracking of floating particles are performed using video data captured in the vertically downward direction with respect to the sonde unit 20 to detect an abnormality. However, it is also possible to detect and track floating particles using video data captured in the horizontal direction with respect to the sonde unit 20 to detect an abnormality.

[0046] In the video data obtained by taking a horizontal shot of the sonde unit 20, for example, if there are cracks in the shaft in the vertical direction near the center of the video and the water in the shaft is flowing out, as shown in Fig. 9(a), the suspended particles will move.

[0047] Therefore, in the same manner as described above, when the weighted corrected normalized velocity spectrum P’(θ,t) for each time t is obtained from the moving velocity vector V(θ,t) of the suspended particles, as shown in Fig. 9(b), peaks appear near θ = ±π / 2.

[0048] Also, in the video data obtained by taking a horizontal shot of the sonde unit 20, for example, if there are cracks in the shaft in the horizontal direction near the center of the video and hot water is flowing in, as shown in Fig. 10(a), the suspended particles will move.

[0049] Therefore, in the same manner as described above, when the weighted corrected normalized velocity spectrum P’(θ,t) for each time t is obtained from the moving velocity vector V(θ,t) of the suspended particles, as shown in Fig. 10(b), peaks appear near θ = 0 and θ = ±π.

[0050] Also, in the video data obtained by taking a horizontal shot of the sonde unit 20, for example, if there are cracks in the shaft in an oblique direction near the right side of the video and the water in the shaft is flowing out, as shown in Fig. 11(a), the suspended particles will move.

[0051] Therefore, in the same manner as described above, when the weighted corrected normalized velocity spectrum P’(θ,t) for each time t is obtained from the moving velocity vector V(θ,t) of the suspended particles, as shown in Fig. 11(b), peaks appear near +π > θ > +π / 2 and 0 > θ > -π / 2. Note that the peak that appears near +π > θ > +π / 2 is more prominent.

[0052] Also, in the video data obtained by horizontally photographing the sonde unit 20, for example, when there are hole-shaped cracks in the shaft near the center of the video and hot water is flowing in, as shown in Fig. 12(a), the suspended particles will move.

[0053] Therefore, in the same manner as described above, when the weighted corrected normalized velocity spectrum P’(θ,t) for each time t is obtained from the moving velocity vector V(θ,t) of the suspended particles, as shown in Fig. 12(b), a flat and high-level spectrum appears over the entire range of θ.

[0054] Thus, even when the weighted corrected normalized velocity spectrum P’(θ,t) for each time t is obtained from the video data obtained by horizontally photographing the sonde unit 20, a spectrum with a shape corresponding to the abnormality of the shaft appears, so that the abnormality can be detected.

[0055] Although the preferred embodiments of the present invention have been described above, the present invention is not limited thereto, and various modifications can be made without departing from the object of the present invention.

Explanation of Signs

[0056] 10 Abnormality detection device 12 Borehole camera 20 Sonde unit 21 Camera 22 Centrallizer 30 Armored cable 42 Winch 100 Shaft 110 Ground part

Claims

1. An abnormal location detection device for detecting an abnormal location in a shaft, comprising: Detecting a predetermined number of floating particles in a predetermined frame and detecting the center position of the shaft from video data captured by a borehole camera connected to the abnormal location detection device, and using the detected predetermined number of floating particles as specific floating particles to detect the specific floating particles in the next frame of the frame, thereby detecting the movement of the specific floating particles and calculating a movement speed vector V(θ, t); For the specific floating particles, calculating a movement speed |V(θ, t)| from the movement direction θ(t) at time t and the movement speed vector V(θ, t); Calculating a descending speed |Vz(t)| of the sonde part of the borehole camera at time t based on the depth Dz(t) of the sonde part of the borehole camera; Calculating a normalized speed vector P(θ, t) of the specific floating particles by scalar division of the movement speed vector V(θ, t) of the specific floating particles by the descending speed |Vz(t)| of the sonde part; Calculating a weighting coefficient f from the center position of the shaft, the position of the specific floating particles, and the normalized speed vector P(θ, t); Calculating a weighted corrected normalized speed spectrum P'(θ, t) for each time t by multiplying the normalized speed vector P(θ, t) of the specific floating particles by the weighting coefficient f; An abnormal location detection device configured to determine whether there is an abnormality in the shaft based on the weighted corrected normalized speed spectrum P'(θ, t) at time t.

2. An abnormal location detection method for detecting an abnormal location in a shaft, comprising: Capturing video data with a borehole camera; Detecting a predetermined number of floating particles in a predetermined frame and detecting the center position of the shaft from the video data, and using the detected predetermined number of floating particles as specific floating particles to detect the specific floating particles in the next frame of the frame, thereby detecting the movement of the specific floating particles and calculating a movement speed vector V(θ, t); For the specific floating particles, calculating a movement speed |V(θ, t)| from the movement direction θ(t) at time t and the movement speed vector V(θ, t); Calculating a descending speed |Vz(t)| of the sonde part of the borehole camera at time t based on the depth Dz(t) of the sonde part of the borehole camera; By scalar-dividing the movement velocity vector V(θ, t) of the specific floating particle by the descent velocity |Vz(t)| of the sonde section, the normalized velocity vector P(θ, t) of the specific floating particle is calculated. From the center position of the shaft, the position of the specific floating particle, and the normalized velocity vector P(θ, t), a weighting coefficient f is calculated. By multiplying the normalized velocity vector P(θ, t) of the specific floating particle by the weighting coefficient f, a weight-corrected normalized velocity spectrum P'(θ, t) for each time t is calculated. An abnormal location detection method for determining whether there is an abnormality in the shaft based on the weight-corrected normalized velocity spectrum P'(θ, t) at time t.

Images