EP2971492A2

EP2971492A2 - Method for determining a position of a water/cement boundary between pipes in a hydrocarbon well

Info

Publication number: EP2971492A2
Application number: EP14710316.2A
Authority: EP
Inventors: Sergey Ivanovich Krivosheev; Evgeni Lvovich Svechnikov; Georgy Petrovich ZHABKO; Andrey Aleksandrovich Belov; Yuri Eduardovich ADAMIAN
Original assignee: FMC Kongsberg Subsea AS
Current assignee: FMC Kongsberg Subsea AS
Priority date: 2013-03-15
Filing date: 2014-03-17
Publication date: 2016-01-20

Abstract

A method and system has been disclosed for determining a position of a water/cement boundary in an annular area between two concentric pipes in a hydrocarbon well. The method comprises the steps of running a well tool (10) into a central pipe of the well, where the well tool (10) comprises a tool housing (11), at least a first and a second pulse generators (14, 34) and at least one signal recorder (16, 17, 18) provided within the housing (11), generating a first and second electromagnetic pulses by means of the first and second pulse generators, respectively, the first and second pulses being separated by a time interval,thereby providing physical vibrations in the central pipe of the well; recording reflected acoustic signals from the well by means of the signal recorder (16); repeating the generating and recording steps with different time intervals of the pulse generators in the well; organizing the recorded signals in a two-dimensional representation; filtering the organized recorded signals in order to identify, in the two-dimensional representation, a hyperbola (B); and providing an apex of the hyperbola (B) as the determined position of the water/cement boundary.

Description

METHOD FOR DETERMINING A POSITION OF A WATER/CEMENT BOUNDARY BETWEEN PIPES IN A HYDROCARBON WELL

FIELD OF THE INVENTION

The present invention relates to a method and system for determining a position of a water/cement boundary in an annular area between two concentric pipes in a hydrocarbon well.

BACKGROUND OF THE INVENTION

Cavities are often filled with a material for insulation or other purposes. In one instance this can for example be a tank with double walls where the cavity between the walls is filled with cement or other hardening material. In another instance it can be a special purpose building, for example a power station having walls where the cavity is filled with cement. Sometimes it may be necessary to ascertain the quality of the filling but where there are difficulties due to inaccessibility or safety reasons. One typical example of such a cavity is the annular space between the casing strings of a hydrocarbon well. A typical hydrocarbon well construction consists of a number of coaxial pipes called casing strings that are successively installed in the well as the drilling progresses. Normally, the first pipe (conductor pipe) is set in the well by being bonded to the surrounding formation with cement that is pumped down the pipe and allowed to flow up in the space between the conductor pipe and the surrounding ground. Then, after drilling further down a second casing normally called surface casing is installed in the well and again the casing is set by filling the annular space between the pipe and the borehole resp. conductor pipe with cement. Then, depending on the length of the hole drilled, and the rock structure, successive casing strings with diminishing diameters are introduced into the borehole and hung off from the wellhead. These casings are normally cemented only partway up from the bottom of the borehole. Lastly, production tubing is installed into the well down to the producing formation and the casings are perforated to allow fluids to enter the well to flow up through the tubing and through the Christmas tree into a flowline. When cementing each pipe the normal practice is to calculate the amount of cement needed, based on the annular space and the length of the space designed to be filled. However, it is often difficult to calculate the exact amount of cement needed and the cement level may be lower than intended. In the case of surface casing it is desirable to fill the annular space all the way up to the mudline (seabed), but this may not always be achieved, leading to so-called cement shortfall. The top of the surface casing may therefore be filled with a fluid (water or brine) instead of cement resulting in that the surface casing string is not bonded to the conductor pipe all the way up to the mudline. In such a case the part of the surface casing that is not cemented can be regarded as a free-standing column that, if subjected to loads, can be damaged.

The surface casing carries a wellhead and is the principal load-carrying structure for the equipment mounted on top of the wellhead. It serves both the purpose of being a foundation for external loads, such as production equipment (Christmas tree) and for borehole support against the formation. A well will be subjected to various loads during its lifetime. In for example a workover situation, a BOP and riser is attached to the Christmas tree, the riser extending to the surface. The movements of the riser and the use of drilling equipment can set up cyclic loads in the wellhead and the surface casing string (See Fig. 1). This may induce fatigue in the casing string.

Another cause of loads comes from the casing strings being subjected to loads from being heated by the producing fluids.

If the cement has filled the annular space completely and, in addition, has bonded properly with the steel pipe cyclic loads will be spread along the length of be casing and transferred to the conductor pipe and the ground. However, if there is a length that has not been properly filled that part of casing can act as a free-standing column (ref. above) and cyclic loads can lead to fatigue and damage of the casing. It is also possible that the point where the top of cement level is can act as a breaking point because of the movements of the column above.

Similarly, heating and cooling of the casing may induce loads that can lead to fatigue problems and deformation of the casing.

As can be understood from the above it is therefore of prime interest to find out if the cement job is properly executed, e.g. the annular space is properly filled. The main purpose of the invention is therefore to find he level of the cement from which the length of the column can be determined.

If later work has to be performed on the well the BOP and riser is reattached to the Christmas tree so that operations can be carried out in a safe manner.

Both during drilling and (if necessary) workover operations the wellhead is subjected to external loads, as explained above. How this affects the wellhead depends on the length of the free standing column. A longer column will be more vulnerable to fatigue. If the length of the free standing column can be determined it can be calculated how much load the wellhead can be subjected to and this will in turn determine how much work that can be done. This enables an operator to predict the operational lifetime of the well and to ensure the integrity of the well structure.

One method for non-destructive logging of layers of different materials comprises the creation of a magnetic pulse within a pipe to cause the pipe to act as an acoustic transmitter. One such example is disclosed in US Patent No. 6595285 where there is described a method and device for emitting radial seismic waves using

electromagnetic induction that generates a magnetic pressure pulse that causes a distortion within a pipe and which utilizes the elastic restoring property of the pipe to cause it to become an acoustic transmitting device. This can be used for generating seismic waves in the subsoil. In US Patent No. 3752257 a similar device is located within a conductor pipe and used to measure acoustic velocity within a formation. The reflected acoustic signals are reflected from the formation and recorded by two receivers and the delta travel time between the receivers is recorded. It is also stated that this apparatus can be used to measure the quality of the cement bond between the conductor pipe and the earth formation. However, there is no further explanation on how this may be achieved and our research has found that this is not a reliable way of determining the cement level.

In both these examples of the known art the transmitter is located such that the acoustic waves only have to traverse one pipe wall, e.g. the conductor pipe. If the device is to be located in a fully completed well there is the challenge to create a signal that is both strong enough to penetrate through several different casing pipes and to be able to distinguish between the reflected signals from the various casings.

In WO 201 1/1 17355 belonging to the applicant, this problem is addressed by using a signal of very short duration. Because of the short duration of the signal it is possible to separate the reflections on a time lapse basis. The speed of the acoustic waves are different in cement (a solid) than in water. When transmitting signals at various points in the well it will be possible to find the spot where the signal is different. This, in theory, marks the exact location of the top of the level of cement. In addition to the problem of separating the various reflections from each other there is also the problem with signal noise. This can be signal noise being generated by the system itself, but also second and third reflections from the various casings. The latter of course becomes even more complicated when the reflected signal comes from an annulus that are several layers away from the receiver, as is the case of the annulus between the conductor and the surface casing, known in the art as the "D" annulus. Both the transmitted and the reflected signal must in this case pass through four casing pipes. There may also be reflected signals travelling along the pipe that also can produce noise.

US-4805156 is directed to an apparatus and method for obtaining azimuthally dependent measurements for evaluating the casing cement bond quality. The apparatus and method employ a plurality of sonic transducers disposed in near contact with the casing in defined segmented locations about or around the exterior surface of a downhole tool. The transducers are arranged in a plurality of arrays, each array comprising four transducers, two transmitting and two receiving transducers, disposed along a single arc about or around the exterior of the tool. The transducers are arranged so that two adjacent transducers in each array perform the same function in the array so that the bond quality between the similarly

functioning and adjacent transducers is determined. The sets of arrays of

transducers provide determinations in a continuous and complete loop around the casing. The casing cement bond quality is determined by measuring the attenuation of the compression wave from a pulsed acoustic wave traveling along the arcs of the helices.

In view of the above background, there is a need for an improved method for determining a position of a water/cement boundary in an annular area between two concentric pipes in a hydrocarbon well.

SUMMARY OF THE INVENTION

The invention relates to a method for determining a position of a water/cement boundary in an annular area between two concentric pipes in a hydrocarbon well, as set forth in the appended independent claim 1.

The invention also relates to a system for determining a position of a water/cement boundary in an annular area between two concentric pipes in a hydrocarbon well, as set forth in the appended independent claim 1 1.

Advantageous embodiments have been set forth in the dependent claims. DETAILED DESCRIPTION

In the following, embodiments of the invention will be described in detail with reference to the enclosed drawings, where:

Fig. 1 is a simplified sketch of a completed well supported by the seabed;

Fig. 2 is a partial illustration of the well of fig. 1 , showing the instrument located in the production tubing;

Fig. 3 is an illustration of a well tool according to the invention;

Fig. 4 is a schematic diagram of an induction coil with its associated circuitry;

Fig. 5 is a schematic view of an induction coil and accompanying field lines;

Fig. 6 is a schematic view of an induction coil according to the present invention; Fig. 7 illustrates a reflected signal from one pulse;

Fig. 8 illustrates a simulation of reflected signals from several pulses fired at different heights in a well; Fig. 9 illustrates theoretical waveforms at the boundary between water/cement;

Fig. 10 illustrates reflected signals from several pulses after processing, where the theoretical waveforms from fig. 9 are overlaid; and

Fig. 1 1 illustrates how the cement level may be deduced from fig. 10.

In fig. 1 there is shown an illustrative embodiment of a completed hydrocarbon well 1. The well is completed with a wellhead 2, production tubing 3, a first intermediate casing 4, a second intermediate casing 5, surface casing 6 and conductor casing 7. The annulus between the surface casing 6 and the conductor 7 is shown filled with cement 8. Cement is normally provided between the drilled hole and the conductor casing, and between the conductor casing and the surface casing. As mentioned above the annular space between the conductor and the surface casing should ideally be filled with cement all the way to the wellhead. The annular spaces between the other casings are normally only filled partway up from the bottom with cement, the amount determined by the formation characteristics. It should be noted that there may be used more than these casings for the foundation of the well, depending on the seabed properties etc. The top end of the production tubing is connected to a tubing hanger that in turn is anchored in the well head or Christmas tree (depending on type of completion) while its lower end is fastened in the first casing with a production packer, as is well known in the art.

In fig. 2 there is shown a part of the well in vertical section showing the casing strings and with the position of the pulse generator 14 and receiver 16 indicated inside the production tubing 3. There are also lines indicating the signals going from the pulse generator and being reflected back to the receiver. In Fig. 3 there is shown a sketch of the well tool 10 intended to be used in a well pipe. The tool 10 comprises a tool housing 1 1 , and a first pulse generator 14 for generating an electromagnetic pulse, which due to the magnetic properties of the pipe will cause the pipe to vibrate. The tool 10 further comprises signal recorder(s)

16, 17, 18 for recording signals representing the vibrations being reflected back from the pipes in the well. Since we are investigating acoustic signals a preferred signal recorder is a hydrophone. The tool may be held in a central position by centralizers (not shown). The first pulse generator 14 and the signal recorder(s) 16,

17, 18 are provided within the housing 1 1 , preferably along a the well tool's longitudinal axis. A second pulse generator 34 is also located within the housing 1 1. It is preferably located within a short distance from and above the first pulse generator 14. The distance between the pulse generator centers may for example be around 120 mm. With an inductor length (height) of about 70 mm this means they are located about 50 mm apart. Other distances may be suitable.

The housing 1 1 is preferably a steel tube that will fit in the well. However, in the region of the pulse generator(s) the tube should be made of a material that is transparent of the magnetic pulses such as glass or plastic. The pulse generators would preferably be held in place in the housing using plates (not shown) that also preferably are made of plastic or Plexiglas.

Each pulse generator 14, 34 that is housed within the tool 10 further comprises a power supply and charging device 22 and a data storage system 24. Further the tool may comprise a cable head 26 for attaching the tool to a cable 30. The cable 30 communicates between the tool and the surface control equipment that may comprise a first control unit 32 for the control of the tool, and a second control unit 33 for receiving and processing data from the tool. An ultrasonic absorber (not shown) may be located between the pulse generators and also between the pulse generator(s) and the signal recorder(s) to avoid unwanted noise in the system. The tool may be coupled to a tractor or similar device for moving the tool in the well.

In Fig. 3 there are shown three signal recorders. However, there may be only one signal recorder located above or below the first pulse generator or there may be one signal recorder located above and one located below the first pulse generator. Each pulse generator 14, 34 comprises an inductor Ls and a power supply device HV, c, which, in use, supplies electrical power to the inductor Ls. Thereby an electromagnetic pulse is generated. The tool and the well pipe are arranged in such a way that the electromagnetic pulse is providing physical vibrations in the well pipe. To this end, the well pipe may be made, at least partly, of a magnetic material. Each inductor may comprise a metallic core, the metallic core may e.g. be a cylinder.

The power supply device HV may be separate for each pulse generator 14, 34.

Alternatively, one common power supply device HV may be used for each pulse generator.

In an aspect, the cross sectional gap area Agap of an annular gap between an outside of the inductor Ls and an inner surface of the well pipe is substantially equal to an inner cross sectional area Ainner of the inductor Ls. In this context, "substantially equal" may, e.g. mean that the ratio between the cross sectional gap area Agap and the inner cross sectional area Ainner is in the range 0.7 to 1.3. More advantageosly, the area ratio may be in the range 0.9 to 1.1 , and even more advantageously, the area atio may be in the range 0.95 to 1.05. Particularly advantageously, the cross sectional gap area Agap is equal to the inner cross sectional area Ainner.

The well tool 10 may advantageously comprise a centralizing device which is configured to positioning the well tool 10 in a central position within the well pipe. The inductor may advantageously have an inductance in the range of 10 * 10^"6 H to 40* 10^~6 H.

The power supply device may advantageously comprise a capacitor, c, connected to the inductor, Ls, wherein the capacitor, c, is configured to discharge its energy over the inductor. Also, the power supply device may comprise a switch, s, connected between the inductor Ls and the capacitor c.

In another aspect the well tool 10 is provided for determining or measuring the presence or absence of cement in an annular area between two concentric pipes in a hydrocarbon well. In such an aspect the well tool comprises a tool housing 1 1 , a firstpulse generator 14 and a second pulse generator 34 provided within the tool housing 1 for generating a magnetic field, where each pulse generator 14, 34 comprises an inductor, Ls, and a power supply device, HV, c, for supplying electrical power to the inductor Ls and thereby providing that an electromagnetic pulse is generated, in such a way that the electromagnetic pulse provides physical vibrations in the pipe being closest to the first pulse generator 14.

In such and aspect the well tool further comprises at least one signal recorder 16 provided within the tool housing 1 1 for recording reflected acoustic signals from the well. Further, a first distance, HI , between the signal recorder 16 and the first pulse generator 14 is substantially equal to a second distance, H2, between the first pulse generator 14 and the annular area. In this context, "substantially equal" may, e.g. mean that the ratio between the first distance HI and the second distance H2 is in the range 0.7 to 1.3. More advantageously, the distance ratio may be in the range 0.9 to 1.1 , and even more advantageously, the distance ratio may be in the range 0.95 to 1.05.

Particularly advantageously, the first distance HI and the second distance H2 are equal.

Advantageously, the well tool 10 may comprise a centralizing device which is configured to positioning the tool 10 in a central position within the well pipe. The second distance H2 may advantageously be measured in a radial direction in relation to the well from the center axis of the inductor Ls and the center of the annular area. The well tool 10 may advantageously be provided in the innermost pipe of the well.

The first pulse generator 14 may e.g. be located at a distance between 10 and 20 cm from the signal recorder 16. Advantageously, the well tool 10 may comprise an ultrasonic absorber 15 located between the pulse generator 14 and the signal recorder 16.

In a particular aspect, the signal recorder may be located above the pulse generators 14, 34. In this particular aspect, a second signal recorder may also be arranged, and in particular, it may be located in close proximity to the first signal recorder.

Additionally, a third signal recorder may also be arranged, and in particular, it may be located below the first pulse generator 14, at the distance substantially equal to or equal to HI below the pulse generator.

In any of the mentioned aspects, the tool 10 may thus comprise signal recorder(s)

16, 17, 18 for recording signals representing the vibrations being reflected back from the pipes in the well. Since acoustic signals are investigated, a preferred signal recorder may be a hydrophone. The tool 10 may be held in a central position by centralizers (not shown). The pulse generator 14 and the signal recorder(s) 16,

17, 18 are provided within the housing 1 1.

The pulse generators 14, 34 are housed within the tool 10 that may further comprise a power supply and charging device 22 and a data storage system 24. Further, the tool may comprise a cable head 26 for attaching the tool to a cable 30. The cable 30 may provide communication between the tool and control equipment, for example surface control equipment that may e.g. comprise a first control unit 32 for the control of the tool, and a second control unit 33 for receiving and processing data from the tool.

An sound absorber (not shown) may be located between the pulse generator 14 and the signal recorder(s) and may be used to prevent acoustic pulses from the inductor to reach the signal recorder and create noise in the system. The tool may be coupled to a tractor 20 or similar device for moving the tool in the well.

In Fig. 3 there are shown three signal recorders. However, there may be only one located above or below the pulse generators or there may be one located above and one located below. In a preferred embodiment there is only one signal recorder which preferably is located above the pulse generators.

The distance between the pulse generators and the signal receiver in relation to the distance to the target may have significant effect. As shown in Fig. 2 the outward waves travels outwards to the D annulus and get reflected back as acoustic waves to the signal recorder. As mentioned above, the distances involved are very small. The standard nominal diameter of a surface casing is 20 inches (50 cm) and a normal size for the conductor casing is 30 inches (75 cm). If we regard the center of the well as the datum, the signals will only have traveled 25 - 35 cm before they reach the surface casing resp. the conductor pipe. In fig. 2, the distance HI between the pulse generator 14 and the closest signal receiver 16 is indicated. Moreover, the distance H2 between the pulse generator 14 and the D annulus is indicated. More specifically, the distance H2 is indicating the horizontal distance between the center axis of the pulse generator 14 and the center of the D annulus. As is known in the art, see for example Fig. 10 in US 6595285, it may be desirable to have a large distance between the pulse generator and the signal recorder. This is no problem when doing seismic surveys since the signals may travel several thousand meters. However, in the confined circumstances in a well and with many scattered 3^rd, 4^th or even higher reflected signals, the separation becomes very important. This technology would not give satisfying results in the confined environment of a well.

The applicant has found that a particularly advantageous result is obtained when the distance HI between the first pulse generator 14 and the signal recorder is substantially equal to, or equal to, the distance H2 between the first pulse generator 14 and the annulus being analyzed.

Based on the abovementioned exemplary dimensioning, that means that the signal recorder should be located about 30 cm from the first pulse generator 14 when the D annulus is analyzed. But a small deviation from this is possible so between 20 and 40 cm will still enable a good separation of reflected signals. In the case of having signal recorders both above and below the first pulse generator 14 they should both be the same distance (HI) from the first pulse generator 14. In the case of having two signal recorders located above the first 14 and second 34 pulse generators (as shown in Fig. 3) they are preferably placed as close to each other as possible.

Arrangements with several signal recorders enables recordings to be compared with each other and can be used to check for anomalies or to find (and eliminate) noise. Another possibility is as use as backup in case of failure.

In fig. 4 there is shown a schematic drawing of a preferred embodiment of a pulse generator. Although the pulse generator in fig. 4 has been indicated by the reference numeral 14, it should be noted that the second pulse generator 34 may have similar or identical structure as the first pulse generator 14. The first pulse generator 14 comprises a charging device, for example a high voltage power supply HV for charging an energy storage device, for example a capacitor C. The capacitor C is connected to a series connection of a switching device S, at least one inductor L and a resistor device R. In fig. 4, the at least one inductor L is represented by a first inductor Ls and a second inductor Li. The second inductor Li is shown only to illustrate self inductance, i.e. internal inductance in the pulse generator 14.

Initially, the switch is turned off. The voltage Uo is applied by the high voltage power supply HV to the capacitor C for charging the capacitor. When fully charged, the switch is turned on, and the capacitor C will discharge by supplying a current I through the inductor MS and the resistor R. The current through the magnetic inductor MS generates the electromagnetic signal pulse which will result in mechanical action on the pipes in the well. These mechanical stress waves are transmitted outwards as acoustic waves which are reflected back to the tool as the waves hit the boundaries.

An illustrative example of the inductor MS is shown in Fig. 5 and 6. The inductor comprises a coil 42 with a number of turns, where the number of turns is

determining the electromagnetic discharge characteristics. A supporting sleeve 43 (shown in fig. 6) may be arranged to support the coil 42 during use and also during production of the coil. When current passes through the inductor Ls it will produce a magnetic field as shown in the figure 5.

The requirements of the elements of the pulse generator 14 will depend on the desired parameters of the generated electromagnetic pulse and the characteristics of the system it is being used in. Inductance (L+MS) results from the magnetic field forming around a current- carrying conductor. Electric current through the conductor creates a magnetic flux proportional to the current. A change in this current creates a corresponding change in magnetic flux which, in turn, by Faraday's law generates an electromotive force (EMF) in the conductor that opposes this change in current. Thus inductors oppose changes in current through them. Inductance is a measure of the amount of EMF generated per unit change in current. For example, an inductor with an inductance of 1 Henry produces an EMF of 1 volt when the current through the inductor changes at the rate of 1 ampere per second. It is this electromotive force that is exploited in the invention. When the inductor is placed within a pipe having magnetic properties, the magnetic pressure from the inductor is converted into a mechanical pressure that sets the pipe in motion, as shown in Fig. 5.

The number of loops, the size of each loop, and the material it is wrapped around may all affect the inductance. An inductor is usually constructed as a coil of conducting material, typically copper wire, wrapped around a core either of air or of ferromagnetic or nonferromagnetic material. When current is delivered through the inductor, magnetic field lines will form around the coil as shown in Fig. 4.

The inductance (in Henry) is presented by the general formula for a type of induction coil called an "air core coil". _ μ₀ΚΝ²Α

" ϊ

• L = inductance in Henry (H)

• μο = permeability of free space = 47Γ x 10 ⁷ H/m

• K = Nagaoka coefficient

• N = number of turns

• A = area of cross-section of the coil in square meters (m²)

• / = length of coil in meters (m)

The present invention may, in an exemplary aspect, use an "air core coil" that does not use a magnetic core made of a ferromagnetic material. The term also refers to coils wound on plastic, ceramic, or other nonmagnetic forms. Air core coils have lower inductance than ferromagnetic core coils. If the coil is not placed into a conductive pipe the field lines inside the inductor will be closer together and therefore the field will be stronger on the inside than outside. This kind of coil directs the magnetic pressure outwards, i.e. the magnetic pressure acts to the inductor extending it in a radial direction.

When the inductor is placed within a conductive screen, e.g. a metal pipe such as tubing the field in the gap between the inductor and pipe will be much stronger than inside the inductor. This effect will depend on the size of the gap and will be strongest when the gap is small. The magnetic pressure then acts to the inductor compressing it in the radial direction.

When the coil is placed within a conductive pipe the general formula can also b expressed thus:

Where

• N = number of turns

• g = gap between coil and pipe

• d = median diameter of coil (see Fig. 6).

•

We have also the following possible parameters:

• do = outer diameter of coil

o this can also be expressed as D-2g where D is inner diameter of pipe

• di = inner diameter of coil, representing the magnetic air gap inside the coil o This can be expressed as D-2(g-w), where 2w is the difference between the outer diameter do and the inner diameter di of the coil • 1 = length of coil in meters Parameters g, d and 1 are exemplary illustrated in fig. 6. In fig. 6, the housing 1 1 of the tool has been removed for clarity and ease of understanding.

The inventors have found that particularly advantageous result for limiting noise in the recorded signals depends on the position of the first inductor Ls and also the size of the inductor Ls in relation to the conductive pipe. This is realized when the cross sectional area of the annular gap area around the coil is equal to the cross sectional area of the inductor inner cross section.

In fig. 6, the cross sectional gap area Agap can be expressed as:

D do

A = *^■(-) - π{—¥

In fig. 6, the cross sectional area Ainner of the inductor inner cross section can be expressed as: di _ _Λ

Ainner = ττ(— }^*"

As described above, particularly advantageous results may be achieved when Agap and Ainner are substantially equal, in the sense that has already been disclosed. Specially advantageous results may be achieved when Agap = Ainner. In this case the field value in the gap is nearly equal to the field inside the inductor. Magnetic pressure will then act on the inductor in the radial direction equally from both sides. In this case the inductor is mechanically balanced and has minimal displacement. This results in minimal inductor acoustic emission and hence less noise in the received signals. Such an exemplary design of the coil is illustrated in Fig. 6. The coil 42 of the first inductor Ls is here placed inside a conductive pipe which in this example is the production tubing 3. The coil may be wound around a supporting sleeve 26 of a non-conductive material. The pipe has an inner diameter D, and the coil has an outer diameter do and it can be readily understood that D - do= 2g where g is the gap between the outer side of the inductor and the inner side of the pipe 3. The length of the inductor is 1.

In an alternative embodiment a conductive (metallic) cylinder is arranged inside the coil. This will function as a balancing element, allowing equalized magnetic pressures inside and outside coil. Due to its mechanical strength it will actually not generate acoustic noise itself. In this case the gap between coil and pipe can be reduced and this may result in lower energy consumption needed for generating of strong enough magnetic field.

In use, the pulse generator is charged up, and when the switch is closed, the inductor will discharge an electromagnetic pulse. The pulse will transmit to the pipe and set the pipe in oscillation. This oscillation excites from the pipe and propagates as pressure pulses through the layers of pipes. As it reaches each layer the pipes will be set in motion and this motion creates acoustic waves that will be reflected back and be recorded by the signal recorder. Several exemplary tests have been performed, using different values and

parameters:

Voltage Uo: 3 - 15 kV

Capacitor C: Capacitance C = 10 - 100* 10^"6 F

Magnetic device MS: Its inductance is L = 10 - 40* 10^~6 H

Initially, the switch is turned off. The voltage Uo is applied over the capacitor C for charging the capacitor until a voltage of 3 - 15 kV is achieved, as mentioned above. The voltage Uo is applied via the wire 12. When fully charged, the switch is turned on, and the capacitor will discharge by supplying a current I through the magnetic device L and the resistor R. During tests, the switch was turned on for periods between 20 - 200 μβ. Even shorter periods of 4 - 20 have also been tested. This short duration is achieved by the geometry of the coil.

The current I will, with the values given above, have an amplitude value in the range of 5 - 20 kA. The current through the magnetic device MS will generate an electromagnetic signal pulse which will result in mechanical oscillations of the pipes in the well. During the tests, the best results were achieved with an energy of the electromagnetic signal of 0, 1 - 3 kJ.

An example of a reflected signal is shown in Fig. 7. As can be seen from this graph the reflected signals coming from the nearest pipe(s) are very strong but get progressively weaker the further away from the signal recorder they are, in the graph this is shown as response time. Therefore, reflections from the area of the "D" annulus are very weak and difficult to interpret.

It was thought that it should be possible to see from the reflected signals whether there was cement or water in the outer annulus, due to the different speed of propagation through these media. However, this has been very difficult to achieve due to the strength of the signals and that the differences we were looking for are relatively small. One possible solution to this problem was to use extremely short duration pulses. The short duration signal pulses result in shorter signals being reflected. Hence, it should be possible to distinguish the reflections from the different structures from each other. Moreover, the distance between the different structures, i.e. the diameter of the different casings, are known. Hence, it is possible to predict when the reflection wave from the different casings will return to the signal recorder, and this information may also be used to analyze the recorded signal. However, this has proved to be difficult due to the reflection from the outer annulus that is obscured by the reflections from the inner layers of the pipe system. To determine where there is cement or where there is water the tool may be positioned at various locations in the tubing. According to such an approach, the tool is positioned at a point in the well a distance below the inferred cement level location. The tool is then moved upwards at small intervals, preferably around 4 cm. At each position the first pulse generator is activated. Each time the first pulse generator is activated a signal of the type shown in Fig. 7 is recorded by the signal recorder. Data representing the acoustic reflections is recorded by means of the signal recorder. The recorded data is transferred to the analyzing device 18 for performing the analysis. The output from the analyzing device is a time-delayed signal that is depicted as lines and curves on a monitor. But for further analysis a two dimensional matrix is used where the columns represent depth of the well and each row represents the time of returned signals. Each element will then show strength of signal. This matrix will be used for the subsequent 2D filtering, as discussed below.

As has been described earlier, in a completed well there may be four or even more layers of pipe between the tool and the target. This means that there are orders of magnitude of reflections from the various pipes. This makes the process of acoustic wave propagation in a system of several concentric pipes very complex. The inventors have found that there are at least three different kinds of propagation. The first kind is waves traveling in the radial direction and reflected by the layers of steel, cement and water as shown in Fig. 2. But in addition there are waves traveling along the pipe in the vertical direction and reflected from the ends of a pipe. Lastly there are waves not belonging to the above mentioned types but are waves that are scattered at various angels and then reflected back to be picked up by the signal recorder. All this means that the differences we are looking for are relatively small. This is because we are looking for the reflection from the outer annulus that is obscured by the reflections from the inner layers of the pipe system.

A production tubing string is made up of lengths of pipe connected together using couplings. At the couplings the tubing has a larger wall thickness than the ordinary pipe. This thicker pipe may result in weaker signals and result in more unwanted noise. In order to avoid the need for varying the vertical position of the well tool, the invention envisages keeping the tool stationary and using sequential firing of two pulse generators to control the strength and direction of the acoustic pulses. The inventors have found that by varying the timing of the sequential activation of the pulse generators we can maximize the signal strength and in this way direct the signal at specific angles. In this way we can effectively make a "sweeping" of the acoustic beam. A good result can be obtained by increasing the time interval between firing in increments of 10 between each sequential activation of the pulse generators. However, shorter increments can be used to obtain a better resolution, for example 5 μβ. The increments may be in the range 5 to 20 μβ. Activating the two pulse generators 14, 34 at the same point of time (0 interval) will propagate the strongest signal in a horizontal plane. Then the interval can be increased by the increment (e.g., 10 μβ), up to a maximum interval. A practical duration of the maximum interval has been found to be about 200 μβ which will result in a maximization of the signal at ± 45°. However, the interval may be in the range 0 μβ to 400 μβ. Advantageous results have been achieved when the interval is varied, in particular incrementally increased, e.g. within the range of 140 μβ to 340

If the upper, first pulse generator 14 is activated first the signal will be directed at a downwards angle. Conversely, by activating the lower, second pulse generator 34 first, the signal will be directed upwards. The activation of the pulse generators in this manner enables us to build up the hyperbola picture and thus determine the cement/water boundary. An advantageous effect of generating a first and second electromagnetic pulses by means of the first and second pulse generators, wherein the pulses are separated by a time interval, thereby providing physical vibrations in the central pipe of the well; recording reflected acoustic signals from the well by means of the signal recorder; and repeating steps the generating and recording steps with different time intervals of the pulse generators in the well, is that an appropriate scan may be achieved without moving the well tool's vertical position.

It should be understood that it is also possible to vary the vertical position of the well tool in addition to using the time delay-type scan approach descibed herein. Hence, a method has been provided for determining a position of a water/cement boundary in an annular area between two concentric pipes in a hydrocarbon well. The method comprises the following steps: First, the method includes a step of running a well tool 10 into a central pipe of the well. The well tool may e.g. be of a type which has been presented earlier in the present disclosure. In particular, the well tool 10 comprises a tool housing 1 1 , at least a first and a second pulse generators 14, 34, and a signal recorder 16 provided within the housing 1 1.

Next, a first and a second electromagnetic pulses are generated by means of the pulse generators 14, 34 included int the well tool housing 1 1. Thereby physical vibrations are provided in the central pipe of the well. The first and second pulses are separated by a time interval.

Next, reflected acoustic signals are recorded from the well by means of the signal recorder 16 included in the well tool housing 1 1.

The two former steps of generating pulses and recording reflected acoustic signals are then repeated for different time intervals of the pulse generators in the well. The repeating may be controlled by the control equipment 32, 33, which includes the first control unit 32 for controlling the well tool 10 and a second control unit 33 for receiving and processing data from the well tool 10. Then the recorded signals are organized in a two-dimensional representation. The organizing may be performed by the control equipment 32, 33.

Further, the organized recorded signals are filtered in order to identify, in the two- dimensional representation, a hyperbola B. The filtering may be performed by the control equipment 32, 33.

Finally, an apex of the hyperbola B is found or identified as the determined position of the water/cement boundary in the annular area between the two concentric pipes in the hydrocarbon well. The determinatio of the apex may be performed by the control equipment 32, 33.

In an aspect the method also includes computing a tip point of the hyperbola B by means of one hyperbola leg and parameters of the well pipe structure, in particular known positions of the two concentric pipes, where the tip point of the hyperbola B represents the water/cement boundary. This may also be performed by the control equipment 32, 33. The two-dimensional representation in which the recorded signals are organized may in an advantageous aspect include a two-dimensional matrix with the relationship between the height, or vertical position, in the well and the time delay of the reflected acoustic signal.

Each element of the matrix may advantageously include a numerical value which represents a signal strength value of the recorded signals. More particularly, the matrix may include column positions that correspond to vertical positions in the well and row positions that correspond to acoustic time delay values.

In any of the above aspects, a filter may be used in the filtering steps to filter out signals with a direction of propagation that is perpendicular to an axis of the well tool. In Fig. 8 there is shown a diagram of reflected signals after having positioned the tool at several locations and thereby representing the recorded signals from the total number of pulses. The vertical lines show the waves coming from the edges, i.e. the pipes. Since we know the strength of the signals, the speed of the acoustic waves and the dimensions of the system we can reliably predict which lines represent which pipe. This will give us a horizontal position of the pipe of interest (the conductor or surface casing). It should be noted that in fig. 8, the signals are from an experimental setup with known cement/water boundary and it was known where the cement was (indicated by dashed line) and where the water was located (indicated by dashed circle). However, as seen in fig. 8, it is not possible to see the difference between the signals representing water from signals representing cement. In fig. 8, the darker lines representing the pipes 4, 5, 6 and 7 from fig. 1 are indicated.

Acoustic signals being reflected from a point source will form a hyperbola with the source being its apex. During our experiments we have seen the same types of signals and further investigation revealed that these signals indeed came from the area forming the top of the cement level between conductor and surface casing, i.e. for example the cement level in the D annulus. This has lead us to conclude that the top of the cement level can be viewed as a discontinuity that can be regarded as a point source. Our investigations has found that at least some of the signals, especially the type 2 and 3 signals, will hit this discontinuity from above and below and that these signals will be received by the signal receiver as a hyperbola.

Theoretically that should enable us to find the cement level by finding the shape of the hyperbola and thereby calculate the apex, and hence the cement level because the apex is in the same horizontal plane as the cement level boundary. However, while we are able to see the signals these types of signals are extremely weak, representing about 2 - 5 % of the total reflections and therefore very difficult to discern from the other signals.

To amplify the signal it is therefore necessary to filter out unwanted signals and possibly also amplify some signals to enable us to find the hyperbola. For this purpose a simulation of the well system is performed. This simulation will return an expected hyperbola based on specific well characteristics. From the simulation we measure the angle of the leg(s) of the hyperbola in relation to the horizontal plane. Then by applying 2D filtering to emphasize signals that lies along the angle returned by the simulation the signals representing the legs of the hyperbola can be amplified and the location of the cement boundary can be found. Filter coefficients of course are specially selected to amplify the feature that we are looking for (a line with a specific inclination). It should be noted that it is normally very difficult to see the area around the apex directly since these signals are usually totally obscured by the signals from the pipe(s), as discussed below.

The theoretical simulation may give us the hyperbolas B with a shape as indicated in fig. 9. The apex of one of the hyperbolas B is also indicated in fig. 9.

Furthermore, S denotes steel and W denotes water. It should be observed that different well designs will return other hyperbola shapes. The signals being reflected from below of the cement level usually are much weaker than those signals being reflected from above. Under certain conditions the signals representing the lower leg of the hyperbola are so weak that they cannot be separated from the noise. And with only one leg known it is not possible to calculate the location of the apex point with desired accuracy. As we have also seen the area around the apex usually is obscured by the much stronger signals coming from the pipes.

However, in these circumstances we can use the information about the distance to the pipes that is obtained from the type 1 signals, as discussed above. From these we can locate the spatial position of the annular area of the cement level as distance from the signal recorder along the vertical (i.e. distance from well center), but not the horizontal position (height). But by knowing one leg of the hyperbola we can by combining the two enable us to find the apex of the hyperbola and thus the location of the cement level boundary.

The picture shown in Fig. 10 represents exemplary signals received by the signal receiver when firing the pulse generators as disclosed, wherein the cement/water boundary is present in the model of the well. Hence, the cement/water boundary is expected to be found. The received signals are placed into a two dimensional matrix where each element is a numerical value representing the strength of the signal. The columns are the vertical position in the well and the rows are the time delay value.

In fig. 10, the theoretical hyperbolas is overlaid on the representation of the signals. It is still not possible to deduce any results from fig. 10, as there is a lot of noise in this picture.

A filter may therefore be used on the signals illustrated in fig. 10 in order to reduce noise. The result of the filtering process is shown in fig. 1 1. In fig. 1 1 , a signal area BA indicated by a dashed circle. This area is found to represent the apex of a part of the upper half section of the hyperbola B from fig. 9. We also know the position of the area representing the space between the two pipes 6 and 7. Hence, by

extrapolating the hyperbola B based on the signal area BA, the hyperbola B can be drawn as shown in fig. 1 1.

The tip of the hyperbola (i.e. the horizontal line through the tip) represents the cement/water boundary.

It is important to note that the amplitude of this reflection very significantly depends on the position of a hydrophone relative to an inductor. This reflection is not very strong but due to its specific properties it can be reliably detected after the appropriate signal processing. Because of the numerous pipes that the waves will travel through there will be both first, second and third order reflections. As can be seen in Figs. 8 and 9 there are very strong reflections coming from waves being reflected from above at an angle. Then there are similar waves being reflected from the bottom. As can be seen in the middle of the figure there are also waves being reflected back at a specific angle and these waves are the indicator of the presence of the cement level.

Claims

1. Method for determining a position of a water/cement boundary in an annular area between two concentric pipes in a hydrocarbon well, comprising the steps of: running a well tool (10) into a central pipe of the well, where the well tool (10) comprises a tool housing (1 1), at least a first and a second pulse generators (14, 34) and at least one signal recorder (16, 17, 18) provided within the housing (1 1); b) generating a first and second electromagnetic pulses by means of the first and second pulse generators, respectively, the first and second pulses being separated by a time interval, thereby providing physical vibrations in the central pipe of the well; c) recording reflected acoustic signals from the well by means of the signal recorder

(16); d) repeating steps b) and c) with different time intervals of the pulse generators; e) organizing the recorded signals in a two-dimensional representation; f) filtering the organized recorded signals in order to identify, in the two- dimensional representation, a hyperbola (B); g) providing an apex of the hyperbola (B) as the determined position of the water/cement boundary.

2. Method according to claim 1 , comprising the step of computing a tip point of the hyperbola (B) by means of one hyperbola leg and known positions of the two concentric pipes, where the tip point of the hyperbola (B) represents the

water/cement boundary.

3. Method according to claim 1 , wherein the two-dimensional representation includes a two-dimensional matrix with the relationship between a height in the well and a time delay of the reflected acoustic signal.

4. Method according to claim 3, wherein each element of the matrix includes a numerical value representing a signal strength value of the recorded signals.

5. Method according to claim 4, wherein the matrix includes column positions corresponding to vertical positions in the well and row positions corresponding to acoustic time delay values.

6. Method according to one of the claims 1-5, wherein, in the filtering step f), a filter is used to filter out signals with a direction of propagation that is

perpendicular to an axis of the well tool.

7. Method according to one of the claims 1-6, wherein the time interval separating the pulses is in the range from 0 to 400 μβ.

8. Method according to claim 7, wherein the time interval separating the pulses is in the range from 140 to 320 μβ.

9. Method according to one of the claims 1-8, wherein the time interval separating the pulses is increased by increments between each sequential activation of the pulse generators, the increments being in the range 5 to 20 μβ.

10. Method according to claim 9, wherein the increments are in the range 5 to 10

1 1. System for determining a position of a water/cement boundary in an annular area between two concentric pipes in a hydrocarbon well, comprising a well tool (10), configured to be run into a central pipe of the well; control equipment (32, 33) communicatively connected to the well tool (10), for controlling the well tool (10) and for receiving and processing data from the well tool (10); where the well tool (10) comprises a tool housing (1 1), at least a first and a second pulse generators (14, 34) and at least one signal recorder (16, 17, 18) provided within the housing (1 1), the first and second pulse generators being configured to generating a first and second electromagnetic pulses, respectively, the first and second pulses being separated by a time interval, thereby providing physical vibrations in the central pipe of the well; the signal recorder (16) being configured to recording reflected acoustic signals from the well; the control equipment (32, 33) being configured to repeating steps b) and c) with different time intervals of the pulse generators in the well; organizing the recorded signals in a two-dimensional representation; filtering the organized recorded signals in order to identify, in the two-dimensional representation, a hyperbola (B); and providing an apex of the hyperbola (B) as the determined position of the water/cement boundary.

12. System according to claim 1 1 , wherein

the control equipment (32, 33) includes a first control unit (32) for controlling the well tool (10) and a second control unit (33) for receiving and processing data from the well tool (10).

13. System according to one of the claims 12-13, wherein

the well tool (10) comprises an ultrasonic absorber.