Patent Grant
11644441
The field relates generally to inspecting a tubular, such as a well or pipe. In particular, the invention relates to methods and devices for imaging surface features using acoustic waves.
Lengths of pipes and wells are commonly inspected to assess their condition. For example, the metal may become corroded, pitted, or thinned. Additionally, the location and state of engineered features such as valves, drilled holes, and perforations may be measured. Such tubular inspection is commonly performed by tools equipped with cameras, ultrasound arrays, x-ray detectors, and magnetic sensors. Ultrasound is particularly useful in hydrocarbon wells where fluids are visually opaque but still pass sound waves.
As these pipes and wells may be several kilometers long, the logging speed of the tool becomes an issue. Such tools may be deployed on coiled tubing, tractor, wireline or pumped down using fluid pressure to move thru the tubular. To obtain high resolution images, the transducer array must capture a frame every few millimeters. However, at practical memory limits, frame rates and processing speeds, it becomes very difficult to log wells that are many kilometers long.
Existing ultrasound tools comprise an array of piezoelectric elements distributed radially around the tool housing. The top surface of each element faces radially away from the tool towards the wall of the tubular. The reflected waves are received by the same elements and the pulse-echo time of the waves are used to deduce the distances to the inner and outer walls and voids therebetween. The elements may be angled slightly off radial, such that some of the energy reflects away from the transducer and some backscatters off features, per PCT Application WO 2016/201583 published Dec. 22, 2016, to Darkvision Technologies. If the angle of incidence is too high (measured from the surface normal), most of the energy is reflected away and little energy penetrates the back off the pipe.
An acoustic wave created by one or more of these elements is focused by a lens and/or through time-delay beamforming, such that energy is generally focused along a line to form a what is called a scanline. The scanlines are then assembled to form one frame that approximates one planer or conical shaped slice of the well. Thus the processor can treat the reflections as resulting from features at that precise spot to achieve a resolution of the spot size. The spot of interest in the next frame will move about one spot size in the direction of logging. These slices are assembled into a 3-D volume in a process called slice stacking.
In accordance with the invention, there is provided a system for inspecting a tubular. The system comprises an image device having an acoustic transducer (preferably operating in the ultrasound band) and an elongate housing defining an axial direction, the transducer arranged to sonify an area of the surface of the tubular with a pulse that intercepts axial locations of said area at different times. A processor is arranged to receive reflected signals from the transducer and to interpret the reflected signals using their time of flight as axial locations of features on the surface of the tubular.
The transducer may comprise a plurality of transducer elements, the device further comprising a circuit to activate a subset of the transducer elements as a phased array to sonify the area and to receive reflections.
The transducer elements may be separated radially around the housing, emitting and receiving phased scan lines at multiple azimuths of the device.
The system may comprise apparatus for moving the device axially through the tubular.
The processor may be further arranged to shift and combine signals from the transducer at a plurality of partly overlapping sonified areas.
The processor may be further arranged to calculate attributes of the reflection signals and identify surface features from those attributes.
The ultrasound transducer may be arranged to emit a defocused beam to the sonified area.
The sonified area may be longer than 5 mm in an axial direction of the tubular, preferably longer than 10 mm.
The ultrasound pulse may sonify said area at an angle of incidence selected between 20 and 70°, measured from a surface normal, preferably selected between 30 and 70°.
The system may comprise a reflector of the device arranged to reflect ultrasound waves from the transducer to the sonified area.
The focal point of the transducer may be one of: radially inward of the transducer; radially inward of the surface, radially outward of the surface, or at infinity.
The transducer may be arranged to rotate about the axial direction of the device and sonify the surface at plural azimuths.
In accordance with the invention, there is provided a method of operating an imaging device comprising an ultrasound transducer to inspect a tubular. The method comprises: deploying and moving the imaging device axially through a tubular; generating an ultrasound pulse with the ultrasound transducer to sonify an area of the surface of the tubular, which pulse intercepts axial locations of the sonified area at different times; and processing reflection signals from the transducer using their time of flight to interpret reflection signals as axial locations of features on the surface of the tubular.
The method may repeat the step of generating an ultrasound pulse to sonify multiple areas that are partially axially overlapping.
The method may shift and combine reflected signals from the multiple sonified areas to remove noise and reinforce reflections from features.
The method may process reflection signals from the multiple sonified areas to calculate registration of features on the surface.
The method may comprise moving the device through the tubular at a speed relative to a frame rate, such that surface features are captured by multiple frames, preferably at least 3 frames, more preferably at least 5 frames.
The method may render the reflection signals as a geometric model of the surface of the tubular and displaying the model to a user.
The method may sonify the area at an incidence angle selected between 20 and 70°, more preferably 30-70°.
The method may sonify a second area of the surface substantially perpendicular to the tubular surface and process the reflection signals from the second area to determine depth data of the tubular.
The method may comprise rotating the transducer to sonify the surface at plural azimuths of the device in a frame.
The method may comprise stitching together radially adjacent reflection signals from each frame to render slices for a geometric model of the tubular.
The method may comprise moving the transducer towards the surface of the tubular, generally adjacent the surface, preferably within 5 centimeters of the surface.
Various objects, features and advantages of the invention will be apparent from the following description of embodiments of the invention, as illustrated in the accompanying drawings. The drawings are not necessarily to scale, emphasis instead being placed upon illustrating the principles of various embodiments of the invention.
Similar reference numerals indicate similar components having the following key:
With reference to the figures, a device and method for imaging the surface of tubular are described. In accordance with one embodiment of the invention, there is provided an imaging device 10 deployable by a wireline 17 at an operations site 18 for imaging features in a tubular 2, illustrated in FIG. 1. The imaging device 10 generally comprises at least one acoustic transducer 12, a deployment system 17, and optionally one or more centralizing elements 20. Tubular 2 may be used to transport or access water, oil or gas. This term is intended to include cased wells, uncased wells, open holes, boreholes, water pipelines, oil pipelines, tubing, and boiler tubes.
The ultrasound transducer may be a single element or comprise a plurality of transducer elements, typically arranged as an array, preferably a radial array. In the simplest case, a single elements sonifies a single spot to provide an image of that area. However, in preferred cases, the single transducer is a spinning transducer that rotates about the longitudinal axis of the device, sonifying multiple areas, sweeping out a helix, as the device moves axially.
In a preferred embodiment of an array, plural elements (i.e. an aperture) cooperate as a phased array to sonify an area. This may be a single area of interest. However preferably, plural elements of the radial array are electrically selected and beam steered to provide multiple scan lines to sonify multiple areas in a frame. As the device moves axially, multiple frames redundantly capture areas to image a large section of the tubular.
The transducer directs acoustic waves, towards the surface at a high angle of incidence such that the majority of the reflected wave is from surface features, rather than depth features. The present device and method are most useful to image a flat surface, at least flat along the direction of the scan line. The optimal system design and operating settings is a tradeoff of logging speed, wavelength, focus, sonified area, frame rate, overlap, and angle of incidence, which may be optimized per job based on the well's diameter, fluid, or surface roughness. In the following discussion, preferred embodiments and ranges are given based on practical parameters, such as Signal to Noise ratio and feature resolution.
The angle of incidence is preferably greater than the critical angle, determined by the speeds of sound for a wave passing from well-fluid to tubular material using Snell's Law. Most of the sound wave will bounce off a metal tubular from water (i.e. and thus will not enter the tubular), at angles of incidence above 30°. At the other end of the range, angles of incidence greater than 70° return too little signal and the surface features are shadowed and occluded by other features. Thus the preferred angle of incidence (measured from the surface normal) may be set at 30 to 70°. Surface features in the reflected waves can be discriminated based on their Time of Flight (TOF), provided that the sampling period of the electronics is shorter than half of the separation in arrival time of pulses. Thus, a higher angle of incidence and higher sampling rate provides a higher longitudinal resolution, but with less signal strength. The inventors have found that a preferable angle of incidence is selected between 30-70°.
In an alternative embodiment, the transducer sonifies the tubular at an incidence angle of 20-30° (i.e. less than the critical angle). Here, although some of the acoustic energy enters the tubular, the processor can separate the internal and surface features using a correlation computation between two successive frames that share some surface features, preferably ignoring reflections from deeper features (i.e. later in time than the initial surface reflections). A high frame rate is desirable here to ensure that features are captured plural times within the small sonified area.
Transducers
The device 10 comprises an acoustic transducer, which transducer is preferably an array of plural transducer elements. An acoustic scan line is generated and received using phase delays for a subset of elements, called the aperture.
The frequency of the acoustic waves generated by the transducer(s) is generally in the range of 200 kHz to 30 MHz, and may be dependent upon several factors, including the fluid types, velocities in the well, and the speed at which the imaging device is moving. In most uses, the wave frequency is ultrasonic, operating between 1 and 10 MHz. Higher frequencies (shorter wavelengths) are able to discriminate finer features on the surface by returning distinct reflections from each feature. This does require greater sampling rate, data processing and data memory. The frequency can be manually or automatically changed based on the velocity of the device and/or fluids in the well.
The number of individual elements in the transducer array affects the resolution of the generated images. Typically, each transducer array is made up of 32 to 2048 elements, preferably 128 to 1024 elements. The use of a relatively large number of elements generates a fine resolution image of the well in the plane of the array. The transducers may be piezoelectric, such as the ceramic material, PZT (lead zirconate titanate). Such transducers and their operation are well known and commonly available. Circuits 15 to drive and capture these arrays are also commonly available.
The number of transducer elements, their pitch, and driving frequency λ affect the imaging quality. In acoustic transducers, a useful rule is that the pitch should be in the range of λ/2 and λ to minimize grating lobes. Thus, in one embodiment, there may be 128 elements, spaced 300 μm (equal to λ) apart, driven at 5 MHz in water.
As shown in
When the array faces the purely radial direction, known as a thickness probe, the reflections indicate the distance to surfaces and defects in the walls. When the radial array is formed on a frustoconical surface, at a cone angle of 30-60°, the wavefront is directed partially toward the longitudinal axis and reflects surface and depth features from a focused spot. As the cone angle β increases, the angle of incidence β/2 increases, the angle from of the longitudinal axis decreases (90°−β/2) and thus the amount of reflections due to surface features increases. There is a corresponding reduction in depth reflections.
In a preferred embodiment of the invention as shown in
In prior systems, the array of acoustic transducers is arranged to focus on a spot of interest in order to reflect back as much energy as possible. The focus may be at the surface of the well or a void therewithin.
In contrast to such systems, the present imaging device does not focus the wave at the surface of the wall but rather spreads the wave energy over an area (sonified area 15 in
As shown in
Thus features near the element return reflections before those further away, and via the TOF conversion, the processor can compute the locations of reflections relative to the sonified area 15. The axial location (relative to the transducer) of a feature i depends on its time-of-flight (TOF), Speed of Sound in the fluid (SoS) and the constant angle of incidence α (typically equal to β/2 when the tool and tubular axes are aligned) by:
Angle of Incidence
The high angle of incidence may be obtained by fixing the face of the transducer at the desired angle (see
While a reflector adds noise from its own surface imperfections, it does provide flexibility in the tool design and type of applications possible. Reflectors may be added to existing tools that currently use a radial facing array. Reflectors of different diameters may be selected on site to enable a tool of a fixed size to image tubulars of various diameters. Conversely a fixed angle transducer will sonify an area much further away on larger diameter pipes, with more attenuation of signal through the extra fluid path.
Moreover, means for moving the reflector may be used to change the sonified area. In one embodiment, the reflector is movable with respect to the transducer array, either translatable longitudinally on the housing or rotatable. During operation, the reflector may move relative to the array from a non-interfering position, to an interfering position between the transducer and wall, to set the incidence angle. Movement of the reflector may set a first angle of incidence and then a second (different) angle of incidence. Either the transducer array or reflector may move to change the relative interference. There may also be a partially interfering position, whereby only part of the reflector reflects the sound at the high angle of incidence and the remaining sound continues undisturbed towards the wall. The reflector may be actuated by actuators in the tool housing slide up/down the housing.
Generating and Capturing Reflections
Sections of the tubular are logged by moving the device through the tubular longitudinally, capturing one frame of plural scan lines, every 5-50 millimeters as it moves. Each frame partially overlaps the sonified area in the previous frame to capture features redundantly. Scan lines comprise the transmission and receiving of an acoustic wave at a given azimuthal angle, which wave may be a sinusoidal pulse or characteristic ‘chirp’. A scan line may be generated by a single transducer element but is preferably generated by plural phased-delayed elements to transmit a coherent wavefront at a desired radial angle in the active focusing plane, typically normal to the array surface. Specialized Ultrasound circuits exist to drive and receive arrays of ultrasound transducers, such as LM96511 from Texas Instruments, illustrated in
In a one-dimensional radial array, the beam cannot be focused longitudinally, so this is called the passive focus plane. This focus is set by the lens or reflector geometry. A frame comprises all the active scan lines that are captured in the frame period, where the number of scan lines may correspond to the number of elements.
For example, in prior radial probe systems, the device may move at 1.5 meters/minute scanning slices of the well at 10 frames per second. Thus each scan line in a frame focusses on a spot of the well 2.5 mm long. Reflections from this spot are from all features at this small spot or behind it, so the processor uses this averaged signal. A given spot is not scanned a second time so the processor cannot take averages of several frames or discriminate sub-portions of the spot. Thus the resolution is limited to the spot size. However, at least the reflections are limited to this narrow spot, with a known location.
Conversely in the present arrangement, reflections correspond to all the features in the much larger, de-focused area 15, and the same wave energy is spread thin meaning that the reflected signals are small compared to the noise. However, here the resolution is mostly limited by the precision for discriminating the reflections' Times Of Flight by the receiving circuit, e.g. the sampling frequency, precision of the timing circuit, and ADCs. In a 50 Mhz sampling circuit, adjacent reflections can be separated in TOF by 20 ns (i.e. 30 um in water) long the angled path. This maps to axial resolution by res=30 um/sin(α), which in the case of a 45-degree transducer corresponds to ˜40 um. However, the resolution is more limited by wavelength of the pulse, where a 5 MHz pulse in water is likely to resolve two features as close as 300 um apart.
The area sonified by the directed wave is large compared to prior art ultrasound devices, and presently captures multiple surface features per frame. The transmitted wavefront may be called divergent or defocused at the sonified area. The area receiving the coherent wavefront is larger than the transducer elevation. The sonified area may be 10-50 mm in the axial direction of the pipe with a radial width which is the circumference of the pipe divided by the number of scan lines. Larger areas envelope more features but spread the energy more.
A curvature of the lens or reflector determines the divergence of the beam. The curvature may be convex, concave or flat, provided that the focal point is not the inner surface of the tubular. The skilled person will appreciate that focusing also depends on the relative speed of sound from the lens material to fluid.
In diverging embodiments, the wavefront may have an effective focal point behind the transducer element, i.e. away from the wall and inside the radial array. When a reflector is used, the focal point of the wavefront will be behind the reflector.
In converging embodiments, the wavefront may have a focal point inward of the surface which then spreads out to sonify a larger area at the surface. Alternatively, the wave may be focused radially outward of the inner surface of the tubular, whereby the area sonified at the inner surface is sufficiently large.
A flat transducer can also sonify a desired area of sufficient size, where the spot size is the transducer elevation amplified by the high angle of incidence. Such an arrangement has a focal point of infinity.
Data Processing
Although the wave is defocused across a longer area (in the longitudinal direction) than prior art focused waves, the present system, after data processing, is capable of higher resolution imaging and feature detection. This is achieved through cross-correlation and coherent summation which removes gaussian noise and reinforces real reflections. In a simplified numerical example: logging speed of device 9 m/min (150 mm/s); frame rate 10 FPS (0.1 s/frame); sonified area 60 mm. Thus the device advances 15 mm/frame and each feature is captured 4 times (60 mm/15 mm/frame).
One advantage of this system is that the large area captured enables a given feature to be observed by the same scan line over multiple frames, where the frames are separated by the inverse of the frame rate. Conversely, Gaussian noises do not repeat in each frame. That is, the system redundantly captures features by partly overlapping the sonified areas in plural subsequent frames. The processor compares two or more frames by shifting the signals in time. The amount of shift is proportional to the logging speed of the tool. The time-shifted signals of plural frames are summed, which has the effect of reinforcing the signals of real reflections that appear in plural frames and cancelling noise that appears in only one frame.
The cross-correlation operator may be applied to B-mode or preferably the raw RF data. Operating on RF data increases the sensitivity
While summing over more frames improves the signal to noise ratio (SNR), it does increase the processing power required. It has been found that combining 3-10 frames balances resolution, SNR and processing power requirements. At four frames overlapping, every frame the features will advance by 25% of the sonified area. Assuming the device is moving at a nearly constant, known speed, the processor can estimate an initial time shift to apply to reflection signals from one frame to the next. From this initial shift estimate, known cross-correlation algorithms can be used to hunt for an optimal shift, where the correlation is maximum.
The shifted and summed (or averaged) signals across multiple consecutive frames may be converted to location data to model the features for a continuous line along the surface. To create a 2D internal surface model, adjacent scan lines are stitched together.
Registration
In prior art systems, location of surface points is estimated from the wireline depth but there is no way to be certain that two subsequent scan lines are immediate physical neighbours. The tool may experience stick/slip motion. In the present system, the processor may use multiple frames to determine and output a registration of features, that is the relative or absolute location of surface features along the tubular.
Multiple features are seen in a single frame. The processor may use trigonometry and Time of Flight conversion to determine the spatial separation of any two features. Moreover, a single feature is seen redundantly in two or more overlapping sonified areas. The processor uses the amount of shift needed to coherently sum frames in order to determine how much the device actually moved between frames or how far offset the features in one frame are from the previous frame. Repeating these calculations, the processor is able to build up the axial registration of all features to create a surface model of the tubular. For completeness, in a radial array, the radial registration of features is known from the radial separation of elements that capture them.
Uncased Wells and Formation
While the device and data processing are suited to generally smooth surfaces found in metal tubulars, they may also be used to image and evaluate uncased wellbores. This is especially useful in bores drilled into tight, hard rock, where the surface features captured include texture from drilling and natural fractures in the formation. The edges of natural fractures return bright ‘glints’ at their edges. The location and quantity of these fractures are processed to evaluate the state of the formation for oil production, before and after hydraulic fracturing.
The surface texture of uncased wells may be indicative of the drilling operation. By processing the pattern of features, the texture may be determined and used to evaluate the quality of the drilling in real-time, known as logging-while-drilling.
Movable Transducer Pad
In larger diameter tubulars, the sonified area on the surface increases for divergent waves and the signal is more attenuated by the increased path through the fluid. In one embodiment, shown in
The device and translation mechanism may be a robot as described in patent application GB2572834 on 16 Oct. 2019, entitled “Device and Method to Position an End Effector in a Well,” which is hereby incorporated herein by reference. The preferred embodiment therein is a Delta configuration for lateral translation with little commensurate axial displacement. The transducer 12 is connected to an end effector connected to one end of and translated by linkages, which are connected at their other end to the body of the device 10. The remaining body of device 10 is not shown in
As described in GB2572834, a further actuator is provided to rotate the end effector and transducer about the longitudinal axis Z, while scanning, such that the entire surface of the tubular is captured.
Yet another of the actuators may extend the transducer axially to controllable interfere with the reflector 22. As described above, the relative position of transducer and reflector may move between a) an interfering position to capture pure surface features and b) a non-interfering position to capture thickness features.
Depth Data
In the embodiments above, the time of flight is used to detect surface features, rather than to acquire depth information (where waves would penetrate the surface and reflect off voids and outer walls). Thus there is very little depth data using purely high incident angled waves.
In an alternative embodiment, some of the transmitted acoustic wave is directed radially towards the wall and reflected back from features at various depths. This may be accomplished by 1) a bi-focal transducer top layer or lens that directs a first part of the wave radially and a second part at the acute angle, 2) a reflector that is partially acoustically transparent and partially reflective or 3) a reflector physically located to interfere with only part of the wave and not interfere with another part of the wave.
The proportion of the wave that should be transmitted radially or acutely depends on the relative strength of the signals required to return useful data. At high angles, the waveform travels a greater distance and experiences less specular reflection, so a higher proportion of the energy is preferably directed acutely than radially.
The radial distance to the inner (and even the outer) wall is sufficiently closer than the nearest point of the acute angle field of view, such that the time of flight of their reflections are distinguishable by the data processor. Thus the first and second reflections will correspond to the inner and outer wall respectively (potentially followed by additional reflections from wall-to-wall reverberations and voids), subsequently followed by surface reflections from features from the nearest to furthest part of the acute field of view. In a tubular of an expected diameter and a fluid of an expected speed of sound, the processor can set time thresholds, to separate radial reflections from acute reflections.
Thus, in these arrangements, depth data is retrieved at the same time as surface data. The processor may create a 3D model by combining these two signals. Since the locations of these two signals are physically offset, the processor offsets them by a corresponding amount of time before combining them.
This processing also helps with registration of the acute angle reflections. The radial wave may be beam focused to a narrower spot on the surface than the acute angle wave to provide a precise location for those reflections. The first reflection will correspond to the surface at the radially-facing location, and therefore the processor can adjust the offsets to find the surface feature most similar to this within the acute reflections.
For example, a dent, crack or perforation may be detected in the radial reflection as a suddenly longer TOF than the surrounding radial reflections (i.e. from prior and later frames). These surface features are given a precise location in the model mapping. As the device moves longitudinally, the acute angle reflections will return glints/strong reflections from the wall of these surface features, albeit with less location precision.
From knowledge of the geometric offset, the frame rate (fps) and speed of the device the processor can estimate the number of frames to offset between a radial scan and the corresponding acute angle scan. The exact number of frames to offset is determined by identifying corresponding surface features, such as the edge of a perforation.
Centralizing Elements
The imaging device 10 may also include one or more centralizing element for keeping the imaging device in the center of the well.
Deployment System
The imaging device includes a deployment system 16 for running the imaging device 10 into the well and removing the device from the well. Generally, the deployment system is a wireline or coiled tubing that may be specifically adapted for these operations. Other deployment systems can also be used, including downhole tractors and service rigs.
Power & Memory System
The imaging system can be powered by an electric cable run from the well surface or by onboard batteries. The data from the imaging system can be conveyed uphole to the well surface through a transmission line for immediate viewing of the images in real-time. The data may also be stored onboard the imaging device for later retrieval in the event of a communication loss. The imaging system may record images continually or it may be triggered manually and/or automatically, such as through the use of movement triggers.
Rendering
The reflection signals may be displayed to a user in their raw signal form, whereby the 2D image is created from pixels separated by the signal's Time of Flight, and wherein pixel brightness is proportional to signal strength.
With further processing, a 2D geometric model may be created by stitching together cross-correlated signals in the axial direction and neighbouring scan lines in the radial direction. The appropriate scaling is provided by converting the TOF to position data. A geometric model represents the surface features spatially (in radial and depth coordinates) and may be stored or displayed in their native polar form or unrolled as a flat surface.
It will be appreciated that data processing may be performed with a processor on the device, at the uphole operations site, or on a remote computer. The term ‘processor’ is intended to include computer processors, cloud processors, microcontrollers, firmware, FPGAs, and electrical circuits that manipulate analogue or digital signals. While it can be convenient to process data as described herein, using software on a general computer, many of the steps could be implemented with purpose-built circuits.
Although the present invention has been described and illustrated with respect to preferred embodiments and preferred uses thereof, it is not to be so limited since modifications and changes can be made therein which are within the full, intended scope of the invention as understood by those skilled in the art.
| Number | Date | Country | Kind |
|---|---|---|---|
| 1901669 | Feb 2019 | GB | national |
| Number | Name | Date | Kind |
|---|---|---|---|
| 5007291 | Walters | Apr 1991 | A |
| 5108693 | Landry | Apr 1992 | A |
| 5581037 | Kwun | Dec 1996 | A |
| 5717169 | Liang | Feb 1998 | A |
| 5899958 | Dowell | May 1999 | A |
| 6813950 | Glascock | Nov 2004 | B2 |
| 7080557 | Adnan | Jul 2006 | B2 |
| 8746070 | Tippit, Jr. | Jun 2014 | B2 |
| 20180011211 | Leonard | Jan 2018 | A1 |
| Number | Date | Country |
|---|---|---|
| 2015050778 | Apr 2015 | WO |
| WO 2016201583 | Dec 2016 | WO |
| Number | Date | Country | |
|---|---|---|---|
| 20200249203 A1 | Aug 2020 | US |
