NL9002378A - Borehole casing thickness determn. method - selects reverberation segment from reflected acoustic signal - Google Patents

Abstract

The thickness of a casing cemented in a borehole passing through an earth formation is determined by a reflection signal. This is obtd. by acoustic examination using an impulse aimed at a radial segment of the casing and formed by sound waves of frequencies selected so as to stimulate thickness resonance in the casing wall. From the reflection signal a reverberation segment is selected representing that occurring between the casing walls. A spectrum signal is generated representing the frequency spectrum of the segment, and means are provided to determined the frequency of components of this signal which contribute to its peak value. From this a signal is produced representing the casing thickness.

Description

Applicant: Société de Prospection Electrique SchLumberger, Paris,

France.

Short designation: Device for determining the thickness of a casing cemented in a borehole.

The invention relates to a device for determining the thickness of a jacket, which is cemented in a borehole penetrating an earth formation from a reflection signal derived from an acoustic examination of the jacket with an acoustic impulse directed at a radial segment of the shell and formed from acoustic waves with frequencies selected to stimulate a thickness resonance within the walls of the shell.

In completing a borehole, a string of auditory tubes or pipes is placed in a bore and cement is forced into the annular space between the auditory tube and the borehole of the hole, mainly for separating oil and gas generating layers from each other and from water-bearing layers.

If the cement fails to separate one zone from the other, pressurized fluids from one zone may be able to move and contaminate another adjacent production zone. In particular, displacement of water gives an undesirable water intrusion into a producing zone and it may therefore become possible that a well be no longer commercially exploitable.

Cement failure can occur in several ways. For example, for some reason it may be a complete absence of cement behind the section of the ear tube, where cement will have to be present. This will be a serious adhesion failure of the cement, leading to rapid contamination between zones, which should be separated from each other.

Another type of cement failure occurs if the cement is present behind the drill pipe, but a small cement-free annular space exists between the cement and the casing. This annular space can be so wide that a hydraulic connection between two zones is made possible, which leads to undesired contamination.

However, such an annular space can also be so thin that the hydraulic safety effect of the cement has been effectively preserved. Such an acceptable small annular space can arise from the technique used to introduce the cement in the first place. For example, the cement is usually introduced under very high pressure, such as is generated by using a heavy drilling slurry to drive the cement plug down and into the annular space to chase the jacket. The resulting pressure within the casing or casing causes slight expansion of the casing and subsequent contraction when the heavy drilling mud is removed. The magnitude of the contraction depends on the pressure and thickness of the casing and tends to result in a small amount. separation and the formation of an annular space between the cement and the jacket. It is important to know whether the cement is performing its function, i.e. whether the cement bond is hydraulically safe.

Techniques have been proposed to determine the quality of the cement bond. In this sense, the term "adhesion", as used herein, is to be understood to encompass both of those cases where the cement actually adheres to the jacket as well as if there is no adhesion, but instead a small annular micro-space which is so it is small that fluid communication between the zones separated by the cement is prevented.

In other words, the term "good adhesion" means that separation of zones by the cement is suitable to prevent fluid displacement between the zones even in the presence of a small annular space. It is therefore desirable that cement evaluation techniques identify such micro-annular spaces as good cement bonds, while annular spaces that are not suitable for separating zones are recognized as hydraulically unsafe or poor bonds.

The quality of the adhesion is also influenced by the thickness of the cement jacket. Good adhesion cannot be guaranteed if the jacket is too thin.

According to the invention, the thickness of the cement jacket can now be determined efficiently if means are present for selecting from the reflection signal a reflection segment which is to a considerable extent representative of acoustic reflections between the jacket wall, means for generating a reflection wall spectrum signal, which is representative of the frequency spectrum of the reflection segment and of means for determining the frequency of components in the spectrum signal, which is representative of the frequency spectrum of the reflection segment and of means for determining the frequency of components in the spectrum signal, which contribute to a peak value thereof and produce a thickness signal representative of the jacket thickness.

The invention will be explained in more detail below with reference to the accompanying figures.

Figure 1 shows a schematic representation of a device for assessing the quality of the cement adhesion and / or the thickness of the jacket in accordance with the invention.

Figure 2 shows a waveform representation of a preferred acoustic implus generated in the device shown in Figure 1.

Figure 3 shows a graph of the frequency spectrum of the acoustic impulse shown in Figure 2.

Figures 4A, 4B and 4C are illustrative waveforms representative of acoustic reflections obtained in an impulse echo survey technique performed in accordance with the invention.

Figure 5 shows an amplitude rendering curve useful in specifying the operating requirements of a transducer which is preferred for use in an acoustic borehole survey in accordance with the invention.

Figures 6A-6C show illustrative spectra of acoustic reflections observed with an acoustic examination device in accordance with the invention.

Figure 7 shows a block diagram of a signal treatment device for appraising cement adhesion in accordance with the invention.

Figure 8 shows a block diagram of another embodiment of a signal treatment device for appraising cement adhesion in accordance with the invention.

Fig. 9 shows a schematic representation of a further cementation appraisal tool in accordance with the invention.

Fig. 10 shows a clocking scheme of a signal operation to be used with an element adhesion appraisal tool of a type as shown in FIG. 9

Fig. 11 shows timing scheme of signals generated in the signal processing means shown in FIG. 10.

Figures 12 and 13 show in top view and partly in section. to use transmitters in a tool as shown in figure 9.

Fig. 1t shows a side view of part of an acoustic research tool using transmitters as shown in figures 12 and 13.

Fig. 15 shows a schematic representation of a device for determining the thickness of a jacket in accordance with the invention.

Fig. 16 shows an amplitude frequency expansion for different spectra obtained with the device shown in FIG. 15.

Fig. 17 shows a block diagram of a signal treatment device for determining the quality of the cement adhesion and the jacket thickness in accordance with the invention.

Fig. 18 shows a block diagram. of a part of a device for determining the jacket thickness in accordance with the invention.

Fig. 19 shows a cross-section of an acoustic tool for re-examination of one. borehole provided with honor. rotating reflector performs the scanning of a borehole.

Figures 1-3 show a system 10 for accustically examining the quality of the cement bonding between a jacket or tube 12 and an annular amount of cement 11 in a borehole 16 formed in an earth formation 18. An acoustic impulse generating tool 20. is suspended within the jacket 12 with a cable (not shown), which is provided with signal paths along which signals for controlling the tool 20 and for monitoring the progression are transferred between a signal processor 21 in the tool 20 and honour. surface-mounted control and handling equipment 22 var. Signals A depth signal, representative of the depth of the tool 20 in the borehole 1k, is derived on a conduit 2k with a conventional depth sensor (not shown) coupled to the cable, with which the tool 20 along the sheath 12 is moved.

The cylindrical casing or housing 12 is shown in partial cross-section as is the annular cement collar 1 ^ surrounding this casing. In the figure, the shape of the borehole 16 is shown to be uniform and the casing is accordingly equidistant from the wall of the borehole. In practice, however, the borehole wall is likely to be irregular with cracks and crevices. As a result, the thickness of the annular cement layer 11 and the spacing between the jacket 12 and the earth formation 18 may vary.

The cement 11 is shown with different bonding states, which are frequently encountered. At the area 16, the cement is shown as to be adhered to the jacket 12, while at 28 an annular micro-space pa 30, which is hydraulically safe, is present In the area 32, the annular space 30 is shown enlarged to a thickness with which separation between zones in the vertical direction can no longer be achieved, while in the area 3 the cement is completely absent. The cement-free areas at 28, 32 and 3 ^ are normally filled with water or a combination of water and mud. These cement conditions do not necessarily occur as shown and are depicted here for the purpose of illustrating the invention. It is sufficient to note that the cement conditions at the areas 26 and 30 should be valued as good sutures, while those at areas 32 and 30 should be perceived as poor.

Furthermore, the jacket 12 is shown with externally corroded segments 33.1 and 33.2 and an internally corroded segment 33.3, where the wall of the jacket is reduced in thickness. Such corrosion may occur in other areas and may be especially harmful if one occurs in an area leading to hydraulic connection between zones, which must remain isolated from each other. The illustrated corroded Segments 33.1-33.3 may act as actual gaps or act as flake segments, which give a rough surface appearance and may even be part of the good parent material The flake segments are saturated by the borehole fluid segments, so that acoustic examination of the good parent material among the scaly segments can still be performed.

The tool 20 fits into the jacket 12, which is normally filled with water or a mixture of water and drilling mud. The tool 20 is held centrally in the jacket 12 with suitable nitrates not shown in detail, which are known per se in the art In the practical application and implementation of the invention, the tool 20 is preferably held parallel to the wall of the casing, although the tool can be moved relative to the central axis of the casing 12. As further illustrated in Figure 1. some compensation for tilt positions, ie if the tool 20 encloses an angle with the centerline of the casing obtained with the system 10.

The tool 20 further includes a transmitter 36 which acts as a pulse transmitter and receiver. In some circumstances, the transmitter and receiver functions. produced by separate devices. The transducer 36 is oriented to direct an acoustic impulse to an acoustic reflector 38 and then through a window ko to a selected radial segment of the jacket 12. The acoustic impulse is partially passed through the jacket 12 and partially enclosed in the sheath 12, reflections occurring in the radial segment and the thickness resonance var. the cloak.

The term "radial segment" as used herein means the segment of the jacket that extends between its walls and a defined radius that extends at least substantially perpendicular to the wall. the mantle extends from the heart of the mantle, surrounds it.

The type of window k0 can vary there. it is preferably made of such material, shaped and so inclined to the direction of movement of the acoustic impulser. from transmitter 36, that the battery returns can propagate with a minimum of attenuation there. source of reflections. The window k-0 may be made of polyurethane, as sold by the Emerson-Cummings Company as CFC-k1 with an accustic velocity of about 1700 meters / second and a density of about

O

1.1 grams / cm. Such a material shows a similar acoustic impedance as a fluid placed in the space between bror. 3d, reflector 38 and window ko for equalizing pressure over window kc.

The fluid with which the space within the tool between the transducer 36 and window ko is filled is preferably selected for low or minimal damping and an acoustic impedance which will not contrast too strongly with that of the borehole fluid of interest. . frequency range An acceptable fluid is, for example, ethylene glycol.

The window 1 + 0 is inclined at an angle Q, which is determined as the angle between the propagation direction of the initial acoustic impulse of transducer 36 and the perpendicular 1 + 1 on the window surface area on which this acoustic impulse is incident. serves to deflect secondary transfers such as 1 + 3.1 in a direction which avoids interference generated by the window. Suitable annular acoustic absorbing surfaces, such as baffle plates 1 + 5, can be used within the tool to trap and absorb acoustic reflections 1 + 3.2 of the inner wall of the window 1 + 0. The size of the angle O can be of the order of 28 to 30 °, as proposed in U.S. Patent 3, 5θ1 + .75δ.

Although the slope of the window may be 1 + 0 in a direction measured relative to the incident beam trajectory as set forth in U.S. Patent 3,501 + .75Ö or 3-501 + .759, the preferred orientation as shown in Figure 1 is in order to allow the use of a larger reflector 38.

The size of the reflector 38 is significant in that the reflector flat surface affects focusing the acoustic energy on the jacket 12 and capturing a sufficient acoustic return for an improved signal to noise ratio.

If the reflectors according to the above-mentioned U.S. patents are enlarged, it is likely that the internal reflections of their windows are intercepted by the reflectors and refocused on the receiver transducer in interference with the desired acoustic returns from the jacket. However, in Figure 1 of the present application, a larger reflector 38 may be used with effective dimensions sufficient to either focus or retain the beam shape of the acoustic energy directed to the jacket 12 and an important acoustic return to the receiver. transmitter 36.

The slope of window 1 + 0 can be clearly distinguished from that used in the structures of the above-mentioned U.S. patents with reference to the orientation of the internal perpendicular 1 + 1 'on the window relative to the point of impact of the acoustic beam along its path of movement D of the reflector 38. If, as shown in Figure 1, the perpendicular 1 + 1 'lies between the beam path D? and the acoustical receiver function of the transducer 36 can be considered positive as the inclination angle and also the angle of incidence. This angle will also be positive if the internal perpendicular lies between the beam sweeping path and a separate acoustic receiver as used in the acoustic borehole apparatus shown in the Russian patent SÜ. 05 · 095 ·

In the case of a window orientation as shown in the aforementioned U.S. patents, the angle of inclination or angle of incidence can be considered negative because the internal perpendicular to the window is on the opposite side of the acoustic beam path and away from the receiver transducer.

With the window slope as shown in figure 1, care must be taken to avoid directing reflections such as b3.2 cv the transducer 36. The angle of inclination must therefore be positive and sufficiently large, however the angle of inclination must not be so great that reflections such as ^ 3 * 2 are not absorbed.or intercepted by the bear baffles i + 5. ;

Part of the acoustic impulse is passed through the jacket 12 and. in turn, is partially reflected by the next interface, which will be cement material in the area 26, while the areas 28. and 32 in the annular space 30 will be more waterborne slurry at area 3¼.

In the exemplary embodiment of Figure 1, the acoustic transducer 36 is selectively positioned so that its effective distance (the travel time for an acoustic impulse) from the jacket 12 is sufficiently long to cause isolation of interference from secondary transmission if the strong acoustic jacket reflection spring becomes partial. reflected from either a window or the transducer 36 back to the jacket 12 to allow for the generation of new reflections and secondary acoustic returns. A desired total spacing D is obtained by substantially placing the transducer .36 at an axial distance D /. of reflector 38, which in turn is located at a distance Ev from the jacket 12 I "* ά.

The total distance D = Di + Dn between transducer 36 and sheath 12 is further selected to be sufficiently long such that the desired acoustic returns, including those due to reflections trapped between the inner and outer walls 13 and 13, respectively. 13 'of the mantle can be observed. Thus, the total distance D is sufficiently long to include these acoustic returns prior to their decay to a small value as a result of leakage into adjacent media. On the other hand, the total distance D is kept sufficiently small to avoid undesired damping by the drilling mud located outside the tool 20 and the fluid inside the tool 20.

In addition to these spacing considerations, it has been found that the distance between the transducer 36 and the reflector 38 affects the sensitivity of the system on tool positions away from a concentric relationship to the center axis hj of the jacket 12. It will be clear that the tool 20 is provided with suitable centering members known in the art and not shown in more detail. Despite the presence of such centering members, a certain displacement of the tool, indicated as an eccentricity distance e between the center line kj of the casing and the centerline bp of the tool arises due to a number of conditions within the sleeve 12. In view of this, the distance is chosen to tolerate a maximum measure of tool eccentricity e.

The optimum distance value further depends on factors such as the effective dimensions of surface 37 of transducer 36, such as its diameter in the case of a disc transducer 36.

For a disk drive with a diameter on the order of about 1 inch for producing a pulse 50 as indicated in Figure 2 with a frequency spectrum as indicated at 52 in Figure 3, the total distance is generally on the order of magnitude between about 2 to about 3 inches.

Thus, a basis for choosing the total distance D is to ensure sufficient time to receive all those acoustic returns, which contribute significantly to an accurate assessment of the quality of the cement bond in the presence of a small jacket-cement ring space. The total distance D must be long enough to allow the portion in the acoustic returns due to poor cement adhesion to be received free of interference.

The acoustic returns include acoustic reflections, which arise as a result of the interaction of the initial acoustic impulse with different media. A first acoustic shell reflection occurs at the interface between the water or the drilling mud within the shell 12 and the inner shell wall 13. This first reflection tends to be consistently the same as the drilling mud consistency, interior shell wall condition, and tilt of the tool 20. Subsequently, acoustic returns occur as a function of reflections from successive media as well as the leakage of acoustic reflections trapped within the mantle.

In the first jacket reflection, the acoustic portion transferred into jacket 12 is now reflected within the jacket wall and 13-13 'and loses energy during each reflection. The energy lost depends on the reflection coefficients r (the reflection coefficient between the fluid within jacket 12 and the jacket) and r ^ (the reflection coefficient between jacket 12 and the next layer, which may be formed by cement as in the area 26 or by var as in the area 32). the shell walls 13-13 'is a function of the shell thickness. Since a jacket of greater thickness tends to cause longer lasting reflections, the total distance D between the jacket and the receiver transducer must progress accordingly; increases.

When a window, which is perpendicular to the direction of movement of the acoustic impulse, as shown in dotted lines at h2 in Figure 1, is used, the mantle reflection and other acoustic returns produce reflections at the interface between the window h2 and the drilling mud within the mantle 12. Such reflections appear as secondary transfers, which are fed back to the mantle for the production of a second mantle reflection with subsequent reflections in the mantle and thus also secondary acoustic returns. These secondary acoustic returns disturb cement appraisal, especially in the case of good cement adhesion, if the formation also has a smooth surface.

In the latter situation, reflections caused by secondary reflections mix with a significant reflection of the formation giving a totally false impression of poor adhesion.

Furthermore, another criterion for determining an acceptable distance between jacket and receiver may include selecting a distance D ~ between a window k'2 and jacket 12 such that secondary acoustic returns fall below a predetermined percentage of their initial skill. Thus, it can be shown that the number of reflections in the steel jacket 12 in such an area is given by the equation:

where x is the percentage part.

The distance can then be represented as given by the equation: D 7 NL C0 r ÜT 'where L is the thickness of the jacket 12, C0 is the speed of sound in the material within the jacket, mainly water, and the speed of sound in the mantle, namely steel, is.

As a numerical example of achieving an acceptable total distance from jacket to receiver, the values for the materials used can be taken in Table 1 below.

TABLE 1

Acoustic impedance density, speed of sound 2 3 z in g / cm sec. p in g / cm C in ft / sec.

water Zc = 1.5 x 105 f> ^ 1 Cc = I9ZO

steel Z1 = b.6 x I0é p = 7.8 C-, = 19, klC

cement Z2 = 7.7 x 10 ^ - 1-96 C0 = 12,000 and Zg = Z0 in case of poor adhesion.

Using these constants, the coefficient of reflection coefficients can be determined as: r0 = 0.937, rG = 0.731 (for good adhesion), rB = ~ 0.937 (for poor adhesion).

The distance from the jacket to the receiver or D ^ can be determined from the above constants and the time constraints (constraints) · For example, if the reflections in the jacket are to drop to about 5 percent of their initial value, the distance may be from about one and a quarter inch to about three inches for a normal rising area of jacket thicknesses L from about 0.2 inches to about 0.6? inch. By relaxation of the final decay value of the sheath reflections, the distance from source to sheath can be reduced, although about 1 inch is probably the lowest possible limit for D ^. Since the greatest jacket thickness is preferably included, the distance from the transducer 36 to either window 1 * 0 or b2 is chosen such that there is no secondary transfer interference over the time interval of interest. If applicable, the distance D ^ is chosen such that secondary reflections attributable to the window do not give signal interference. If the tool 20 has a window such as the window Uo, secondary reflections from such a window are no longer a consideration in choosing the distances from the transducer to the jacket.

In the choice of the transducer 36, a disc transducer with a diameter to wavelength ratio greater than the unit is used. In practice, a disc transducer with a diameter of about one inch has been found useful. The transmitter pulse is formed with such duration and frequency that a selected radial segment of the sheath on which the impulse falls and a thickness resonance is stimulated. Acoustic energy is transferred into the sheath and bounces back in honor. increased manner with the duration and magnitude of reflections highly sensitive to the material layer near the outer surface of the jacket 12.

However, such sensitivity will not include hydraulically safe annular microspaces such as region 28.

In the choice var. the frequency spectrum of the acoustic impulse var. transducer 36 is a first base determined by the fundamental thickness. fesonanti frequency of sheath 12. Such a resonance allows for a trapping method that traps increased acoustic energy in the sheath. The subsequent reduction of trapped energy in the sheath may be considered as a result of leakage due to the degree of acoustic coupling to neighboring media. The frequency spectrum of the acoustic impulse will preferably include either the fundamental or a higher harmonic thereof. Expressed in mathematical terms, the stimulation frequency in the acoustic impulse is given by:.

f0 = ȣ i 21, where C-j is the jacket compression rate and L the jacket thickness is measured perpendicular to the jacket wall and N is an integer.

An "upper" limit of the frequency spectrum of the acoustic impulse is posed by practical considerations such as mantle roughness, grain size in the metal mantle, and "drilling mud attenuation. Furthermore, the hydraulically safe annular micro-space must appear permeable.

In practical cement bonding applications, "an annular jacketed cement space equal to or less than 0.005 inch (0.127 mm) provides good cement adhesion and thus prevents hydraulic connections between zones, which must be separated from each other. If annular spaces greater than this value occur, they must be considered and established as poor cement attachments. Furthermore, as long as an annular space in thickness is less than about 1/30 of a wavelength of an acoustic wave moving in water, such an annular space is effectively transmissive for an acoustic wave of an equal wavelength. Thus, in terms of annular space between shell and cement, the frequency spectrum of the acoustic impulse can be chosen such that: fo k. Co _ (hole) x 30, where Cp is the speed of sound in water and pat the thickness of the annular is space.

Jacket thicknesses L normally encountered are from about 0.2 inch (5.08 mm) to about 0.65 inch (16s51 mm). As a result, with an effective frequency of from about 300 KHz to about 600 KHz for the acoustic impulse, jacket 12 can be stimulated in a capture mode insensitive to hydraulically safe annular microdynamics.

This frequency spectrum is chosen so that the capture mode can be stimulated with either the fundamental frequency or its second harmonic for the thicker mantles.

Within such a frequency spectrum, the duration of the reflections within the steel jackets becomes sensitive to both good and bad annular microspaces. For an acceptable annular microsphere, the sheath reflections (and their observed leakage) decay faster than for particularly large annular microspaces .

Thus, the acoustic transmitter pulse is formed with characteristics as shown in FIGS. 2 and 3. The transmitter pulse 50 shown in FIG. 2 shows a highly damped acoustic pulse of a duration of the order of about 8 microseconds. The frequency spectrum of such Pulse 50 is shown in Figure 3 with a frequency amplitude curve 52 showing a 6 dB (half power) bandwidth ranging from about 275 KHz to about 625 KHz peaking at about K25 KHz. The thick shrouds, which capture mode below 275 KHz are primarily driven in resonance with a higher harmonic, such as the second, which occurs with significant amplitude in the bandwidth of spectrum 52.

The transmitter 36 can be formed by any number of well-known materials for producing pulse 50 with the frequency spectrum 52. For example, an electrical signal having these characteristics can be generated and amplified to drive a suitable piezoelectric transducer 36 capable of to work as a transmitter and receiver.

Preferably, the transducer 36 is constructed with a piezoelectric disc crystal, which is supported with a critically matched impedance such that an acoustic pulse is generated at the resonance frequency of the disc. The support material has an impedance selected to match with that of the crystal while strongly damping the acoustic impulse to avoid reflections from the support. In some applications, a protective front layer, which is integrally applied to the front face, may be used. the transducer 36. Such a pre-layer is preferably made of a weakly damping material with an acoustic impedance which is approximately the geometric mean between the crystal impedance there. the expected impedance of the borehole fluid. Such a pre-layer has a quarter wave length thickness measured at the center resonance frequency of the crystal.

Since the disc is critically tuned, the acoustic output pulse has a wide frequency bandwidth. Energization of a similar transmitter 36 can then be obtained with an electrical pulse of very short duration, for example, a pulse having a rise time of about 10 to about 100 nanoseconds and a fall time of 0.5 tct about 5 molar eaves can be used.

In the transmit mode, the transducer 36 can be operated in a repetitive manner with an impulse value of, for example, about the order of 100 pulses per second. A circumferential area can be used with such a vessel. about the shroud 12 are scanned as the tool 20 is moved up the shroud by supporting reflector 38 and the associated window 1 * 0 as shown for rotation in the direction of arrow 53.

Figure 5 determines the performance criteria for a suitable transducer 36. The transducer has a mid-acoustic frequency at approximately 1 * 25 KHz with a 6 dB bandwidth of 300 KHz. Figure 5 shows an acceptable received amplitude-response curve 55, if transducer 36 is energized with an impulse drive signal of about 2 microseconds duration and directed to a water / air interface spaced from the transmitter equal to about 100 microseconds of 2-way acoustic wave propagation time, T 1. The output signal from transducer 36 as a result of the interface echo will preferably have an appearance as shown, where the first echo formed from the three main peaks 57 · 1.57 · 2 and 57 · 3 will not have a greater total duration Tp are then about 6 micro-seconds. The level var. the noise immediately after the first echo will be about 50 db below the level of peaks 57 and have a duration of less than about 30 microseconds. The noise level A ^ following interval will preferably be at least 6o db below level A ^ of the peaks are 57.

The control and circuit necessary for igniting the transducer may originate from overhead equipment or from an appropriate clock source contained in tool 20. In both cases, repetitive synchronization pulses are generated on a line 5 ^ (Figure 1) for activating an iropulse network 56 to drive a transducer 36 while simultaneously protecting an input 60 on an amplifier 62 with a signal line 61+.

The transducer 36 responds to the impulse of network 56 with an acoustic impulse of the type shown in FIGS. 2 and 3. The acoustic impulse is directed to the reflector 38 which acts to direct the acoustic energy to the wall of the jacket 12. The effect of the reflector 38 contributes to compensate for variations in alignment of the acoustic impulse from the plane perpendicular to the jacket wall. The reflector 38 may be a flat surface at an angle ** of about 1 + 5 ° to the acoustic energy of transducer 36, or a slightly concave or convex surface.

When the acoustic impulse 50 impinges on the jacket 12, some of the energy is reflected and part is transferred into the jacket 12. The reflected energy is fed back to the transmitter 36 via the reflector 38 and is reproduced as an electrical signal and applied to input 60 from amplifier 62.

The energy transferred into the jacket 12 bounces back which in turn causes further acoustic returns to transmitter 36.

The resulting received output from the transducer 36 may have the appearance as shown with reflection signal waveforms éb, 66 and 68 5η FIG.

The initial segment 70 of each reflection signal waveform gives. the strong initial mantle reflection, the duration of which is in order. of about 5 microseconds. The remaining portion 72 is characterized as a reflection segment because it represents a large number of cycles of impulses representative of acoustic reflections whose magnitudes are about. The decay period varies as a function of the type of cen adhesion, as can be seen from waveforms 6h, 66,68 obtained with respective differently sized annular spaces 30 about the jacket 12.

Except for the initial shell reflection segment 70, the reflection signals 6h, 66 and 63 do not have a highly predictable pattern in which the peaks can be accurately determined and separated. Accordingly, a known technique for comparing neighboring peaks to determine decay time constants for the waveforms is difficult to fit.

Instead, the signal treatment segment 21 of the device 10 acts on each reflection signal by separating the reflecting segment 72 from the initially strong acoustic jacket reflecting segment 70 and then integrating the reflecting segment 72 over a period of time to determine the energy therein.

In the exemplary embodiment of FIG. 1, the reflection signals from the transducer 36 are amplified in amplifier 62, the output of which is applied to a full wave rectifier 76 to produce a DC signal representative of the amplitude of the received acoustic signal at line 7P. wave. The DC signals are filtered in a filter 80 for line 82 of a signal representative of the envelope of the waveforms of transducer 36.

The envelope signal on line 82 is applied to a threshold detector 8¼ which initiates subsequent signal processing by sensing the onset of the initial mantle reflection segment 70 (see FIG. 1). The amplitude at which the threshold detector 8k is operated. can be varied with a selector fitted on line 86 and can be set automatically,

The output of threshold detector 8k on line 88 is used to activate a release pulse on output 90 of an impulse generating network 92. The impulse of this network 92 is chosen to be such that the envelope segment on line 82 is attributable to the initial shell reflection 70 is gated through an amplifier 91 as a shell reflection signal.

The duration of the release pulse at output 90 is selectable such that. the entire mantle reflection segment 70 can be selected in case its duration varies. The DC shape of the mantle reflection signal is applied to an integrator network or peak amplitude detector 96 to generate a signal representative of the amplitude of the mantle reflection 70 on line 98. stored as with a sample and holding network 100 actuated by an appropriate pulse derived on line 102 from network 92 at the end of the pulse on line 90.

The output 88 from the threshold detector 8 ^ is also applied to a reverberation segment selection network 103 having a delay 1θΗ, which generates a release pulse on pulse generating network 106 at a time after the initial shell reflection 70 has ended.

The network 106 generates a segment selection pulse on line 18 starting at the start of the bounce segment 72 and of a duration sufficient to gate the entire envelope shape of the bounce segment 72 (see Figure k) through gates from amplifier 110 to integrator 112.

The segment selection pulse on line 108 begins after the initial sheath reflection and ends after the desired number of acoustic returns of interest are received, but before secondary transmission interference occurs. A usual pulse will be approximately 6 microseconds after the initial sheath reflection is observed, begin and will last for a period of about microseconds after an acoustic impulse delivered as shown in FIGS. 2 and 3 and with a distance D of the order of about 3 inches.

The integrator 112 integrates the envelope shape for a period of time determined by the pulse on line 108. At the end of this last pulse, a signal from pulse generator 106 activates on line 11. a sample and holding network 116 for storing a signal representative of the energy in the reflection segment 72. "

The outputs of sample and retaining network .100,116 are applied to a combiner network in the form of a divider 118, which forms a quotient by dividing the signal representative of the energy in the reflection segment 70 by the normalization signal indicative of the amplitude of the sheath refraction 70 to generate a normalized energy bond signal on output + output line 120, The normalized and very ie signal training 120 can be transferred to courts above the ground to capture the reflection energy as a function of depth on an expander 122. The normalized energy signal may also be applied to a comparator 12¾ for comparison with a reference signal on line 126 derived from a network 128 and representative of the threshold between good and bad cement bonds. output 130 v & n '& e vc equation device 12¾ indicates the presence or absence are of good cement adhesion to and may also be "recorded" on Layer 122 as a function of depth.

With the signal operation execution modes, hell becomes; adhesion signal on line 120 made less susceptible to tool tilting and fluid damping directing the energetic energy through jacket 12 along a plane oblique to the centerline of jacket 12. the received acoustic returns decrease in amplitude and can be interpreted as good cement tea joints, when in fact the cement adhesion may be poor. By using the amplitude of the initial magnitude reflection as a measure of tool tilt and condition of the suspension the bonding signal on line 120 provides a reliable indication of the quality of cement bonding.

In some cases, there may be a need for the acquirer of an adhesion signal that is not normalized or that can be normalized at a later time. In such a case, the output 117 of the sample and holding network 116 is the adhesion signal which is equipment located above ground may be surrendered for recording as such on a tape recorder or a recording device 122 or in the memory of a signal processing device 130 after conversion to a digital form.

After an adhesion signal is generated and a new synchronization pulse occurs on line 5 ^, the synchronization pulse is applied to various reset inputs of sample and holding network 100,116 and integrator 96,112. The reset of the sample and holding networks 110,116 can be delayed for more gradual output until the outputs of integrators 96,112 are ready for sampling.

The selection of a signal representative of the acoustic reflection return 72 is obtained with an impulse generated on line 108 as can be determined with a segment selection network 132. This network controls the length of the delay 1 (A and the width of the release pulse from pulse member 106. As already described above with reference to Figures 1 * A, 1fB and itC, the reflection segment 72 is selected in such a way that the sheath reflection 70 is effectively excluded.

This exclusion can be effectively obtained by the signal processing device 21 as it is activated by the observation of the strong jacket reflections 70, as sensed by threshold detector 8U. The resulting integration of the remaining envelope gives a sharp distinction between a signal of good adhesion and a signal of poor adhesion. For example, the integration of the reflecting segment 72.1 of the waveforms 6b in Figure Ua will be greater than the integration of the reflecting segment 72.3 of the waveform 68 in Figure by a factor of about 3. If the area of the envelopes is compared for an example as plotted in Table 1 with the resulting reflection coefficients for r ^ and r ^ for good and bad cemer.t adhesions, an integration ratio of about 3.8 to 1 between bad and good signals occurs. As a result, a particularly sharp good to poor adhesion contrast, which is likely i is even obtained in the presence of a dense drilling mud within the jacket 12.

With certain types of cement one can wish. an annular microsphere with a thickness on the order of about 0.010 inches (0.25 mm) as good cement adhesion can be seen. In such a case, the frequency spectrum 52 of the acoustic impulse 50 can be adjusted to examine the cement. For example, it is possible to use types of acoustic pulses of different frequency spectrum, one of which has the fundamental frequency and the other which has a harmonic. If the results of these impulses do not give the same reading, it can be concluded that a hydraulically safe annular micro-space is present.

Theoretically, an adhesion will appear as good for an annular micro-space with a thickness on the order of previous wavelength (about -0.08 inches). In practice, the formation of such a large annular space is unlikely and other conventional cement quality assay techniques can be used to identify such improbably large annular spaces as poor cement adhesion.

Figures 6A-6C show the efficiency of the tool 10 when a frequency spectrum is made on the observed full acoustic returns as shown in Figures 1A-1C. The spectra 1-0 of Figures 1A-6c show resp. poor adhesion with a large annular space, an intermediate bonding situation such as with an 0.005 inch annular space and good cement adhesion. The spectra Ho, if originally obtained, may have varied in absolute size because the reflective changes in tool eccentricity e and the coupling of acoustic energy until the cement 11 behind the jacket 12 varies. Thus, for good cement adhesion, the absolute amplitude of the acoustic returns is lower than for poor cement adhesion.

The relative size of dips 11 + 2 varies, however with a larger dip for poor cement adhesion and a smaller dip 1 ^ 2 for good cement adhesion. For the sake of efficiency, the spectra 1A0 are shown with at least in substantially equal amplitudes, so that the respective ips 11 + 2 can be appraised by honor. visual comparison with each other. The meaning of the dips 1.1 + 2 must progress. determined in light of the total energy spectrum of the reflection signal.

The sharp dips ί3 + 2 in the spectra 1 ^ 0 are centered in the capture mode or thickness resonance of the mantle from which the reflections came. In the spectra 1 ^ 0 the dips 1 ^ 2 occur at 0 »5 MHz (500 KHz). ) for a sheath 0.23 inch thick and resemble the effect of a narrow hand width energy trap filter.In case of poor adhesion, such as for the spectrum 1H0.1 in Figure 6a, the dip 1 ^ 2.1 is deep as indication that a relatively significant amount of energy has captured the thickness resonance within the mantle walls 13-13 '.

The improvement in cement adhesion is clearly demonstrated in spectrum "\ k0.2 by a correspondingly smaller amount of energy trapped or trapped within the mantle walls 13-13". As a result, dip 1 ^ 2.2 in Figure 6b is smaller compared to dip 1 ^ 2.1 in figure 6a, while dip 11 * 2.3 in figure 6C is the smallest for good cement adhesion.

Figure 7 shows an exemplary embodiment 150 for appraising the cement adhesion using the sharpness of the dips 1U2 in spectra 1L0 of Figures 6A-6c. The output 63 of amplifier 62 in network 21 is applied to two transmission hand filters 152 and 15 ^ *. The filter 152 is a pass-through hand filter tuned to the jacket thickness resonance frequency of the jacket 12 under aural examination. The pass-through hand for filter 152 is preferably narrow with sharply rising and falling slopes. However, the filter 152 must be sufficiently wide in its frequency band to overlap the frequency range of dips 11 + 2 for the expected tolerance variations in shell thickness. In general, a filter 152 with a transmission hand of about 10% to 15 $ of the center frequency will suffice, although a narrower transmission hand of about 5% may provide a dipamplude indication on lead 156. Both a digital and an analog filter 152 can be used.

The filter 15 is preferably tuned to a separate non-overlapping segment of the spectrum of the signal on line 63 to obtain a reference signal on line 118, which gives an indication of the amplitude of the spectrum of the signal on line 63. Other devices can be used to derive such a reference signal, such as a peak sensing technique described with reference to the embodiment shown in Figure I. The dip amplitude signal on line I56 is normalized thereon by dividing this signal by the reference on line 158 by part network I60. A normalized dip value signal is then available on the output 162 of the divider 160 to obtain an indication of the segment adhesion quality prior to recording or expansion as appropriate.

Figure 8 shows another embodiment for determining the cement adhesion. The output of a transducer 36 on a line 63 of amplifier 62 (see Figure 1) is applied to a high speed analog to digital (a / d) converter 172, which is moved a certain time after an acoustic pulse. This produces a digitized reflection signal formed from successive numerical samples representative of the amplitude of the reflection signal.

The A / D converter can be deactivated for a certain period of time following the generation of an acoustic impulse.

The A / D converter I72 is located downhole in the perf cap 10 and is capable of operating at a very high speed and equipped with a sufficient storage capacity for initial storage and subsequent transfer from the samples at a slower rate to a surface-mounted signal-enhancement device 17 '. The latter device may, if desired, also be contained in the tool 10, but this will depend on the extent of the processing which the signal-processing device 17 ^ must execute.

The sampled digital reflection signal is stored in a memory 176, which may be a solid state memory or a magnetic memory. The memory 176 may be an integral part of the device 17 for immediate processing of the samples or an edge device, which is stored on a later time is approached after examining the borehole 16.

The signal processing device 17 ^ may be programmed to select at 178 those reflection samples Ac, which are representative of all mantle reflections 70 (see Figure 1 *). The procedure may be similar to that shown in analog form in Figure 1. Thus, the reflection samples are scanned. for observing the first sample, which exceeds a certain threshold, and this first sample becomes the arrival time of the mantle reflection. The number of samples following this first sample are then chosen as representative of the mantle reflections 7C (see Figure 4).

A number of sample reflections A, which follow the mantle reflection samples A, are chosen at 180 as representative of the reflection segment 72 in the reflection signal (see Figure h).

Integration of the reflection samples is done by adding the absolute values of the samples at 182. This summing step can be performed if the reflection samples are chosen at 180. However, for the sake of clarity, the summing operation is indicated as a separate step. integrated sum E is stored r.

Integration of the mantle reflection samples A is obtained at 18¾ by summing the absolute sample values and storing the result E.

c

A normalized bonding value CB, representative of the cement bonding quality, can then be obtained at 186 by dividing the integral Er by the integral Ec at 186. The bonding value CB can be stored in a memory or plotted, as desired, at 188.

Figure 9 shows another embodiment for examining the quality of the cement bond. A tool 210 suspended from a cable 211 is equipped with a plurality of transducers 36 arranged along the structure of the tool 210 to provide adequate cement bonding appraisal. circumferential solution. The transducers 36 are axially spaced to allow the large number to be recorded. A practical number of transducers 36 may be ac, where these eight transducers are arranged circumferentially with - • io rooms of 45. The axial distance is selected according to the size of the transducer 36.

Figures 10 and 11 relate to a signal handler 215 for operating a tool 210 shown in Figure 9. The signal handler 215 will be described herein in the form usable for a tool 210 with eight transmitters 36. However, a larger number of transmitters may be included. The signal processing device 215 has an adjustable clock 212 at the output 21¾ of which pulses 216 (see Figure 11) have a value chosen to determine the solution of the cement adhesion test. The clock source can be derived from above-ground devices or any suitable oscillator included in tool 210.

The clock pulses 216 are applied by a delay network 218 to a transmitter selector 220 and a transmitter pulse multiplier 222.

The transducer selector 220 provides a separate outgoing enable signal on line 22b to sequentially identify each individual transducer 36. Subsequently, the multiplexer 222 is released for sequentially firing pulse members 226 coupled to the transducers 36.

The transducers 36 also function as receivers and produce signals on amplification output lines 228 in pre-amplifiers 230, which are operatively associated with each transducer. 36. The output of amplifiers 230 are connected to a receiver multiplexer 232, which is controlled by the transducer identifying signals on line 22b of transducer selector 220. In addition, a segment selection network 23 ^ is activated with each transducer ignition to generate enable signals. 236 (Fig. 11) on an output 238 for effectively clearing multiplexer 232 for selecting the desired segment of the transducer outputs while rejecting or erasing var. the initial channel segments. The output 2 ^ 0 of multiplexer 232 will have an appearance as shown at 2Ui in Figure 11. small noise signal 2 ^ 2 precedes reflection signal 2'Mp, which has an appearance at least in hocflsaak as spring, given in Figures 1 * A-kC.

As further indicated in Figure 10, the reflections cj the output line 2l + 0 were amplified by two variable voltage amplifiers (VGA) 2 ^ 6,218. The amplifier 2ho has its gain controlled by honor. signal on line 2ho and derived from either above ground equipment to adjust for drilling mud damping effects or from an automatic gain control located downhole. The gain of the second amplifier 2k8 is automatically controlled in tool 210 to correct tool off-center position 2 "0, as will be further explained below.

The output 250 of amplifier 2h8 is rectified in a network 76.1 and applied to a sheath reflector receiver network obtained from a gated amplifier 91, integrator 96 and sample and retainer receiver 1CO, as described. the hand of figure 1.

The output on line 250 from amplifier 2 ". 8 is further amplified! in an amplifier 252 with an amount sufficient to compensate enter it. approximated difference in signal amplitude between the mantle refraction. the acoustic returns that are indicative of successive returns. Acceptable compensation may be a gain of about 20 db for amplifier 252, reflections of interest are then applied to a full wave rectifier 76.2 for subsequent integration with devices as described with reference to Figure 1.

Control over the gate amplifiers 9 ^ S110 is generally derived as described with reference to Figure 1 with a threshold sensing device 8b, which responds to the output on line 78 of full wave detector 76.1. A reference threshold value is derived on line 86 as a result of a similar preliminary cementitious bonding test has been made with the appropriate transferor as will be further explained below.

The output 88 of threshold network 8Π is applied to the set input of a latch network 256. Network 25 ^ has a reset input 258, which responds to clock pulses on line 21b (for the delay of network 218). If the threshold detector detects a signal on the line 78, which is greater than the reference value on line 86, a pulse is applied to network 256 which is thereafter prevented from responding to further inputs from the drum detector until network 256 is reset by a pulse on line 21b. The output on line 260 will last as shown by pulses 262 (Figure 11) with an active state on the occurrence of the large mantle reflection.

The integration times and (see also Fig. 11) for signals representative of the sheath reflection and the reflections are derived with impulse networks 92 resp. 1θ6 whose outputs 90, 108 are applied to free gate amplifiers 91, 110.The duration and occurrence of the integration periods and are resp. about 8 microseconds for the mantle reflection and about 30 microseconds below for the reflections.

Subsequent integration of the jacket reflection signal by integrator 96 and the reflecting segment by integrator 112 are terminated at the end of pulses and if the output of amplifiers 91, 110 returns to zero. The outputs of the integrators are sampled at the end of pulse and made the samples available for further treatment with a suitable multiplexer 266 for transferring the samples to the "above ground equipment. Transfer of information can be done using" analog to digital oazette 2 $ 7 and appropriate telemetry equipment 269 for transmission along cable 2h, The integrators 965112 are reset by pulses on line 219 and the sample and Retention net sales by pulses on line 2lk of the transfer logic 271 at the time of clock pulses 2 \\.

As mentioned above, the gain control for amplifier 2148 is automated by sensing the peak value of the jacket reflection on lead 178 with a peak detector 270. The peak value is then converted to a digital value with an A / D converter 272 and this value is placed in a storage network 27 in a location associated with the transducer from which the reflection was obtained.

The next time this transducer is energized, the transducer selector 220 provides a suitable address signal to a read-in-read logic network 275 for applying the previously stored peak value to an gain control network 276 and a threshold reference signal generating network 278.

For gain control, the digital peak value is converted to an analog signal and an appropriate bias is applied to control the gain of amplifier 2lt6.0. Similarly, the threshold reference value on lead 86 is maintained at the appropriate level for each transducer 1 · 6.

The techniques used in appraising the cement adhesion as described above effectively allow an accurate measurement of the eccentricity of the tool as the tool moves along the casing. This technique, as shown in Figure 10, includes a timer 280, which times is energized when a transducer 36 is initially triggered. The timer 280 was deactivated to store a measured time interval if a mantle reflection is sensed by the threshold detector 81+ as shown by the signal on line 260, The measured time intervals for the different transmitters must be the same and any difference can then be attributed to an off-center position of the tool.

The timing network output can be recorded or turned off and appropriately treated for measuring and determining the eccentricity of the tool 210.

The vertical dissolution of the tool 210 is a function of the repetition rate at which the transducers 36 are energized and produce observable sheath reflections and reflections. A repetition rate as high as 100 per second can be recorded to produce a solution as small as about every one- tenths of an inch if the tool is moved along the casing at an examination speed of about 1D inch per second. A signal on line 213 (see Figure 9), representative of the depth of the tool 210, is obtained to release a signal handler 215 to adjust for the difference in level of transferors 36.

Figures 12, 13 and 11+ show a multi-use acoustic power source and detector 300 on a tool as shown and described with reference to Figure 9-The detector / source 300 is mounted radially on a cylindrical housing 302 with a mounting bracket 30 of a central opening 306 for receiving a cylindrical or disc transducer 36. The mounting bracket 30 extends beyond the radiating surface 37 of the transducer with an opening wall 308 slightly expanding outward.

The bracket 30 ^ may be mounted directly on the housing 302, as shown in Figure 12, or with an intermediate spacer 310, as shown in Figure 13. In the assembly of Figure 12, the spacing D between the transducer and jacket may accommodate smaller jackets, for example, about 5 1/2 inches in diameter. The arrangement of Figure 13 can accommodate larger mantles.

Preferably, the radial orientation of the transducers 36 does not include a window or intermediate materials. Furthermore, the distance D between the transducing surface 37 and the jacket 12 is kept as small as possible.

Since too small a spacing D allows secondary transfers to conflict with the reflection of interest, the distance D should not be too small. On the other hand, if the distance D is too great, damping effects of the drilling mud can also become too great. As a result, a compromise regarding distance D must be chosen based on expected damping.

The damping may vary depending on the type of drilling mud used. For example, a heavy or dense drilling mud may cause undesirably high damping. Consequently, when selecting an acceptable gap D, it may be necessary to specify an upper limit for the density of this drilling mud. With such an upper limit, the maximum attenuation can be about 4 to 5 db per inch as opposed to a heavy drilling mud attenuation of about 8 to 10 db per inch.

With these general limitations, an acceptable distance D can be on the order of about one to about two inches for most jackets.

The described arrangement for the tool 20 with a rotatable reflector 38 can be varied in a number of ways. For example, in some instances, it may be desirable to place the reflector 38 in a pad near the wall of the jacket 12 to reduce the damping effects of a dense drilling mud. Care must be taken to ensure that the reflector 38 remains at a sufficient distance from the wall of the jacket 12.

The shell thickness is measured by analyzing the frequency spectrum of the reverberation segment 72 (Fig. 4a) which is representative of acoustic returns due to reflections between the shell walls 13 - 13 '. If an acoustic pulse such as 50 is directed to the sheath 12 traps a significant amount of acoustic energy at the resonant frequency within the sheath walls.

The reflection segment 72 has predominant components in a frequency portion 320 (Fig. 6A-6C) at least substantially in frequency in line with dips 142. The dips 142 increase in depth as the quality of the cement adhesion decreases, but the amount of energy collected between the shell walls increase with poorer adhesion between the cement and the shell. As a result, the actual amplitude of the acoustic returns in the frequency portion 320 will vary. Generally, your actual amplitude of the acoustic reflections within the frequency portion 320 decreases as the acoustic coupling between shell 12 and cement 14 becomes more efficient, that is, as the cement adhesion improves.

This is shown in the spectrum image of Figure 16 with curves 322 and 324 which, respectively. represent the frequency spectrum of a frequency portion 320 for a sledite cement bond and good cement bond.

If thin spots arise in the sheath 12, such as at 33.1 and 33.2 in Fig. 15, it is likely that they affect the appraisal of the cement adhesion. The effect of such thin spots on cement adhesion is not easy to predict and is likely to be a function of such factors as size and cerentium. For example, there is no cement adhesion behind the thin spot 33.1, but since the shell is considerably thinner here, less acoustic energy is trapped within the shell walls 13 - 13 'than in the case of normal thickness, so that the thin spot 33.1 can appear as a good attachment. On the other hand, if an insulated externally thin place, such as 33.2 occurs at a well-adhered area, it may appear that the jacket 12 is poorly adhered. It is therefore advantageous to be able to correlate a measurement of jacket thickness with an appraisal of the cement adhesion to eliminate ambiguities.

i

The measurement of the jacket thickness is made in the device 326 shown in Figure 15 by forming a frequency spectrum of the reflection segment derived on line 63 of Figure 1. The frequency spectrum is characterized by one or more peaks, the largest of which occurs at a fundamental frequency whose wavelength is twice the thickness of the shell. Other peaks occur at frequencies that carry a whole multiple relationship to the fundamental frequency.

Fig. 16 shows different frequency spectra 322, 324 of different reflection segments 72 selected from different signals. It will be clear that, in the representation of the different spectra in Fig. 1, it is not the intention to set an amplitude ratio between the spectrum 52 of the acoustic pulse 50 (Figs. 2 and 3) and the other spectra 322, 324; rather, it is only intended to represent a frequency ratio in that the spectra 322, 324 occur within the frequency bandwidth of the incident acoustic pulse. In practice, the absolute amplitudes of the acoustic spectra will be quite small compared to that of the transmitted pulse.

Of particular interest is the relatif frequency / shuffle between spectra 328 and 330. The frequency difference between peaks 328 and 330 can be attributed to a change in the thickness L of sheath 12. Consequently, by determining the frequency of the peaks, which are mainly due to acoustic returns of the reflections between the shell walls, an indication of the thickness of the shell are obtained.

The sheath thickness L can be derived from the following equation: 1 -!: - h: P · where f ^ is the frequency of the peak in the spectrum, C is the compression rate in the sheath 12 and $ is an integer depending on whether the measured peak for the fundamental frequency (N = 1) or higher harmonic js.

Considering that the frequency spectrum 52 of the acoustic pulse 50 has a bandwidth of about 300 to 600 Khz for use with shells 12 over a wide range of thicknesses, from about 0.2 inch to about 0.75 inch. probably, the second harmonic (K = 2) produces the largest peak in the reflection spectra for the thicker mantles, while N = 1 for the thinner mantles. The value for f 'can therefore be determined prior to an acoustic examination from a knowledge of the jacket type provided in the borehole.

For example, if it is known that an applied jacket has a nominal thickness of 0.362 inches, its fundamental thickness resonance occurs at about 331 Khz for a value of C of 20,000 ft / sec.

As actually measured, the spectrum 322 can show a peak 328 at a frequency fg of about 348 KHz corresponding to an actual jacket thickness of 0.345 inch in a radial segment of the jacket. Spectrum 324 shows a peak 330 at a frequency f of about 303 KHz corresponding to an actual jacket thickness of 0.395 inch.

These measurements demonstrate the solution of the technique by observing a jacket thickness variation of approximately + TL due to fabrication of the face value of 0.362 inches.

In the one shown in fig.15. The gauge thickness is measured by selecting the reverberation segment 72 on a line 332 with a selection network 334 coupled to the reflection signal on line 63. The selection network 334 uses a jacket reflection detector 336 to obtain an impulse at output 338 the leading edge of which is representative of the start of the shell reflection 70 (fig. 4). Detector 336 may be constituted by a rapid response threshold detector 84 or, as shown in FIG. 1, by a full wave rectifier 76, filter 80 and threshold detector 84.

The pulse on lead 338 is delayed by a delay 340 for a period of time matched to the duration of the strong initial jacket reflection 70 to then initiate a pulse feed 342. This impulse network produces a reflective selection pulse on lead 344 to release an analog port 346 for a duration corresponding to the time required to select the portion of the reflection signal that is substantially representative of reflections within the jacketed walls.

A spectrum analyzer 384 responds to the reverberation segment on line 332 to obtain on line 350 a signal representative of the amplitude Λ of the frequency components in the reverberation segment 72, while the output line 352 carries a corresponding frequency signal f representative of the frequency of the amplitude components. on line 350.

The amplitude and frequency signals on lines 350 and 352 are applied separately to anologue to digital converters 354 and 35b which, in a memory 358 of a signal processor 360, the digital signals representative of the amplitude A, <-n the frequency f ^ of the frequency spectrum of the reflection segment 72 and store it.

The operation of the spectrum analyzer 348 and the A / 1 converters 354 and 356 is initiated by the reflection segment selection pulse generated on line 344 through impulse network 342. During this last pulse, a local oscillator is internally converted to spectrum analyzer 348, repeatedly swept through a frequency range to generate the amplitude spectrum on lead 350. Each time the local oscillator is swept through its frequency range, the spectrum analyzer 348 generates a spectrum field of amplitude A ^. and frequency f ^ signals. As a result, during the selection of a single reflection segment 72, a number of spectrum fields are generated and stored in memory 358.

For a non-repetitive reflection segment 72, a separate multiple time base of the local oscillator in the spectrum analyzer 348 may be sufficient to derive an indication of the frequency spectrum. The A / D converters 354,356 are of such a type that an appropriate number of conversions can be made during each sweep of the local oscillator.

Once the spectrum fields formed from frequency and amplitude A..signals are stored in memory 358, the signal hardener 360 is activated to search for a peak amplitude value A at 362. This is done by examining all stored amplitude values A and comparing each value with the subsequent amplitude value and withholding the greatest amplitude value for the following equation. By maintaining the frequency value associated with each retained amplitude value, the frequency f of the peak A ^ can be found and both are conveniently stored at 364.

In some cases, multiple peaks may occur in the stored spectrum samples. Although the largest peak is used to derive a thickness determination, one can also apply both peaks to this and choose that jacket thickness measurement as the correct measurement, which is closest. the nominal value is located.

The observed peak values, both amplitude A ^ and frequency f, can then be recorded, for example on a scribe 122. The frequency f ^ can be recorded directly as an indication proportional to the jacket thickness L or the latter can be calculated on the basis of the above given ratio and then be recorded. Other information may be simultaneously recorded on the recording device 122, such as the well depth on line 24, the detection signal on line 120, azimuth of a rotary scanning reflector on line 37 to indicate the depth and circumferential location of the radial ridge segment, whose thickness was measured.

In an alternative exemplary embodiment for determining the jacket thickness, as shown in Figure 17, the entire reflection signal on line 63 is digitized as described with reference to Figure 8 for cement bond appraisal. The digitization process is initiated upon detection of the mantle reflection arrival by detector 336, which is described with reference to FIG. 15.

The output pulse on line 338 of the detector 336 is a pulse of sufficient duration to allow digitization of an entire reflection signal such as 64 (FIG. 4A). This pulse activates a network 370, which generates a pulse on line 372, with a duration generally about equal to that of the shell reflective segment 70 shown in FIG. 4. The pulse on line 327, in turn, closes an analogous shell logic 374 for this period of time for passing the shell reflection segment 70 to A / D converter 172. The latter digitizes the shell reflection segment 70 and stores the samples in a suitable memory, not shown.

If the grain count reflection segment has continued, the pulse on line 372 goes inactive, which in turn activates a network 342 to obtain a release pulse on line 344 to allow an analog analog bounce gate 346 to conduct a bounce segment 72 through an amplifier 376 with a gain control input 374, to A / D converter 172.

Amplifier 376 allows amplification of the normally weak bounce segment 72 for more accurate signal handling. The digitized reflection signal can be processed downhole or passed up the cable with a suitable telemetry device 380.

A signal processing device 382 is arranged to act on the digitized reflection signal from A / D converter 172. The treating device 382 provides a jacket thickness determination at 384 and a cement adhesion appraisal signal CB at 386.

The jacket thickness is determined by selecting the reflection samples P at step 388 and generating a spectrum thereof at 380 with a Fourier conversion. The spectrum is formed from amplitude values A. and related frequency values F.,

The spectrum is scanned to choose the maximum peak value. This can be done by setting at 392 a counter equal to the number of DN of reflection samples, a constant K = 1 and the values of AMAX and FMAX equal to zero.

A test was made at 394 to determine whether the amplitude value A for the sample K is greater than AMAX. If so then the values for AMAX and FMAX are made equal to A (K) and F (K) at 392. The following samples can then be examined by increasing K and decreasing the counter over one at 398 and examining for which the counter is equal to zero at 400.

If not all samples have been scanned, the counter is not equal to zero and the examination for a maximum spectrum value is rehr.c.ld at 394. Once all samples have been scanned, the mazimal values AMAX and FMAX can be recorded at 384 or the jacket thickness L derived from the formula: L = N 2 (FMAX) *

A cement adhesion appraisal can be made effective by signal treating device 382 using the steps described with reference to Figure 8.

The cement adhesion signal GB varies as a function var. the jacket thickness. This variation can be significantly removed from the cen adhesion signal at 402. This includes dividing the cell adhesion signal CB by a jacket thickness signal L as determined at 4ΓΛ from the frequency measurement FMAX using the jacket thickness ratio as set forth above. This normalization of the cement adhesion signal removes variations due to direct proportional jacket thickness changes, leaving smaller second order jacket thickness effects. Thus, the cement adhesion for a given radial segment can be advantageously appraised in a manner that is substantially insensitive to the jacket thickness in the same radial segment. Cement adhesion normalization to jacket thickness can also be performed directly with a cement adhesive signal available at 182 in Figure 7 or on lead 117 in Figure 1 for normalization by the jacket reflection signal. The latter signal can then be used to further normalize the cement adhesion appraisal as described.

Fig. 18 shows an alternative embodiment for deriving the frequency of a peak in the spectrum of a reflection segment 72. The outputs 350 and 352 of spectrum analyzer 348 (Fig.

15) are recorded on continuous tracks 410.1,410.2 of a storage medium 412 such as a magnetic disk or drum. After recording the output of analyzer 348 for a bounce segment 72, the information is played back for analysis for a coherent signal processing network 414 for sensing, storing and recording the amplitude and frequency peak values A and f.

P P

The spectrum analyzer outputs 350 and 352 are shown to be coupled by logic amplifiers 416,418 to capture / playback heads 420, 422 operatively arranged with respect to magnetic storage disk 412. The amplifiers 416,418 are released by the segment select pulse on lead. 344 (Fig. 15). The amplitude A and frequency f signals are recorded on separate continuous tracks 410.1 and 410.2, which have a sufficient recording length to record an entire reflection segment 72.

After recording the reverberation segment, logic replay amplifiers 424 and 426 are released due to the removal of the disable pulse effect on lead 344 by converter 428. This then plays back the previously recorded amplitude A and frequency t signals - possible.

A peak detector 430 is provided to sense the peak value in the amplitude signals played back by the amplifier 42Λ. The observed peak value is then applied to a comparator 432 along with another replay of the previously recorded amplitude signals on track 410 · 1 to determine the frequency f on it. P

time that the peak occurs.

If comparator 432 recognizes equality between its inputs, a pulse is generated on output line 434 to activate a sample and hold network 436 coupled to sample the playback frequency signal f from amplifier 426. The frequency fp of the amplitude peak value is then stored and made available on output line 438 for recording and use as an indication of the thickness of the sheath 22, as described above.

The recording, peak scanning and peak frequency selection are performed in sequence under the control signals on line 1 * 1 + 0 of a logic control network 1 + 1 * 2. This network is initiated by the impulse on line 3¼ and then by the playback of one recording of such impulses derived from a control track 1 * 10.3 on the magnetic storage medium 1 * 12.

Figure 19 shows another embodiment 1 * 6θ for an acoustic tool for examining the cement hitch and casing, in which, as in Figure 1, a rotating reflector 38 is used. The tool! * 60 is equipped with a stationary transmitter 36 and a longitudinal cylinder 1 * 62, which is arranged centrally and rotatably with respect to the tool 1 * 6o about an axis of rotation 1 * 6! *, which in this exemplary embodiment preferably coincides with the central axis of the tool.

The tool l * 6o has an annular acoustically transparent window 1 * 66, which is arranged between an upper tool section 1 + 68 and a lower tool section 1 * 70.The cylinder 1 * 62 internally bridges the window 1 * 66 and is rotatable in engagement with the top and bottom sections Λ68 and 1 * 70 through armies 1 * 72. .

The. cylinder 1 * 62 has a tubular section 1 * 7! * in which transducers 36 protrude through an open end at 1 + 76. The tubular section 1 + 7! + ends at reflector. 3.8 from which the cylinder 1 * 62 is preferably solid up to its bottom end 1 * 76.

Cylinder 1 + 62 is equipped with a pair of annular radial, protruding flanges 1 * 78.1 and 1 * 78.2. Bearings 1 * 72 are clamped against flanges 1 * 78 with annular bushes 1 + 80 attached to tool sections 1 * 68 and 1 * 70 using bolts 1 * 82. Bearings 1 * 72 fit into axially open annular. grooves 1 * 81 *, 1 * 86 in flanges 1 + 78 resp. bushes I + 80. Bearings 1 * 61 * provide both low friction support in radial direction and push direction. Additional bearings and flanges can be used if required.

The cylinder 1 * 62 has a sturdy strong construction for reinforcing the lower tool section 1 * 70 to which a load generating device, such as an externally mounted centralizer, may be provided.

The cylinder 1 * 62 thus serves as a solid reinforced bridge over the acoustic window 1 + 66. The possibility to use a centralizing element under the rotating reflector 38 allows an accurate placement of the axis of rotation 1 + 61 + with respect to the jacket 12. possible and thus provides an accurate distance from the reflector 38 from the jacket 12.

The acoustic reflector 38 has a reflection angle of a size necessary to allow acoustic connection through a lateral opening '' -90 in the tubular section 1 + 71 +. In front of the opening and subsequently extending with the outer wall of the upper tool section U68, the acoustic window 1 + 66 is formed of a material that has a predetermined acoustic impedance and is equipped with a shape chosen to minimize unwanted acoustic reflection.

The acoustic window 1 + 66 is formed from a material whose acoustic impedance closely matches the acoustic impedance of a fluid, as described with reference to Figure 1, and which is placed in the space between the source 36, reflector 38 and window 1 + 66.The acoustic temperature and pressure coefficients, i.e. the change in acoustic impedance as a function of temperature and pressure for both the fluid and the window 1 + 66 are chosen as close as practicable as possible The acoustic window 1 + 66 may be made of a material as described with respect to the window 1 + 0 in Figure 1 or from polysulfone, a material sold by the Union Carbide Corporation under the trade name Radel, which material has an acoustic velocity of about 2200 meters / second. Therefore, if an acoustic pulse is generated from the source 36 to the reflector 38, the acoustic energy travels through the fluid / window interface 1 + 92 with minimal reflection.

In order to further reduce the effect of acoustic reflections from a window placed between the source 36 and the jacket 12, the window is conically shaped with an angle of inclination de to the reflector 38 as described with reference to Figure 1 in order to use allow a larger reflector 38 and also bend secondary transfers away from the jacket 12.

The transducer 36 in Figure 19 is mounted on a support 1 + 91 + j which is attached to the wall of tool section 1 + 68. An electrical cable 1 + 96 connects the transducer 36 to an electrical circuit (not shown).

A rotary drive for the cylinder b62 is obtained by an electric motor Ιρδ, άίθ is driven within the tool ^ βθ and is equipped with an output shaft 500. A gear transmission 502 connects the motor shaft 500 to the cylinder hè2.

The gear transmission 502 can take many different shapes and is shown for illustrative purposes as consisting of a pair of pinions 50t and 506, the pinion 506 being mounted on a shaft 508 mounted in a bush 510 on the support 1 + 9 ^ -A bevel gear with bevel gears 512,51¾ is used for coupling the shaft 508 and the cylinder k62.

With a tool U60, as shown in Figure 19, the structural integrity of the tool extends below the annular window h66, providing additional strength under the window and allowing its centralization relative to the jacket 12 with a centralizing device. can. be made sufficiently strong for such bending forces as through them. can be exercised from the rotatable cylinder 162.

After the above discussion of techniques for examining a casing cemented in a borehole for appraising cement adhesion and casing thickness, the advantages of the invention will be apparent.

Variations can be made to the above-described embodiments, and the embodiments described and illustrated above are shown for illustrative purposes only.

Claims (7)

An apparatus for determining the thickness of a jacket, which is cemented in an earth formation penetrating borehole from a reflection signal derived from an acoustic examination of the jacket with an acoustic impulse directed at a radial segment of the jacket and formed from acoustic waves with frequencies selected to stimulate a thickness resonance within the walls of the jacket, characterized by means for selecting a reflection segment from the reflection signal which is significantly representative of acoustic reflections between the jacket walls, means for generating a spectrum signal which is representative of the frequency spectrum of the reverberation segment and means for determining the frequency of components in the spectrum signal, contributing to a peak value thereof, and producing a thickness signal representative of jacket thickness. The device according to claim 1, characterized in that the reflection signal is formed from digital samples and the spectrum generating means is provided with means for generating a fourier conversion of samples representative of the reflection segment as the spectrum signal. Apparatus according to claim 1 or 2, characterized in that the means for determining the peak value further comprises means for producing samples of the spectrum signal with associated values of the frequency of the samples and means for scanning of the spit drum samples for a peak value thereof and choosing the associated frequency value as an indication of the thickness of the shell. An apparatus according to any one of the preceding claims, characterized in that the selector further comprises means responsive to the reflection signal for sensing a signal therein representative of an initial acoustic reflection signal from the jacket and means responsive. on the directional mantle reflection signal for selecting that portion that follows the initial mantle reflection. Device according to any one of the preceding claims, characterized by means for generating a strongly damped acoustic pulse from within the jacket in a radial direction to a radial segment of the jacket, the acoustic pulse being generated with acoustic wave frequencies in a bandwidth which is selected to stimulate an acoustic resonance between the walls of the jacket with acoustic reflections and to obtain a reflection signal representative of acoustic returns caused by the acoustic impulse and means for generating digital samples of the reflection signal, wherein the selector includes means for selecting samples, which are representative of the sheath reflections and occurrence following samples, which are representative of an initial sheath reflection, while the generating means is provided with means for generating a spectrum of the selected feedback measurement samples and forming amplitude samples with associated frequency values and the determining means comprises means for determining a maximum amplitude sample and the associated frequency value as an indication of the thickness of the shell. The device according to claim 5, characterized by means for adding the absolute value of the selected samples representative of the reflections in the jacket as a measure of the quality of the adhesion between the jacket and the cement. An apparatus according to any preceding claim characterized by means for selecting samples representative of an initial acoustic mantle reflection of the inner wall of the mantle, means for adding the absolute values of the samples representative of the initial acoustic jacket reflection, means for summing the absolute values of the selected samples, which are representative of the jacket reflections as a measure of the quality of the adhesion between the jacket and the cement and means for forming a quotient generated between the corresponding sums by the summing means for normalizing the measurement of cement bonding quality.