GB2297170A - Branched conductor identification - Google Patents

Abstract

In order to identify a branched conductor 10 such as an underground cable, signals are applied to the conductor at a plurality of points on the conductor, e.g end points. The units 14, 15 ,16 which apply the signals each generate a signal of two components. The first component is of a known magnitude, and the second component identifies the unit which applied that signal. When the signals are detected at a point (e.g. A or B) on the conductor, the detector can identify the signals received, on the basis of the second component. By summing the first components and comparing the result with the predicted magnitude of their sum (since their magnitudes are known), it is possible to determine if the correct conductor has been identified. The signals may be applied periodically, with timings determined by a synchronising voltage pulse. The first signal components may all be of the same magnitude or any one or more may be reduced by a known amount if the unit is unable to generate a signal of that predetermined magnitude.

Description

CONDUCTOR IDENTIFICATION The present invention relates to the identification of a conductor, such as a cable or pipe buried underground.

The prolification of networks of buried cables for many different utilities (electricity, telecommunications, etc) has made it increasingly difficult to identify whether any particular cable, located by a suitable location technique, belongs to a particular utility or not.

Existing identification arrangements fall into two types. Firstly, optical identification techniques involve excavation of the ground in the vicinity of a buried cable, until the cable is visible, and thus can be identified. In order for such an excavation to be carried out, the cable must first be located (i.e. its position determined to enable excavation to take place at the right location), and the excavation thus needed is timeconsuming. Furthermore, excavation involves a risk of damage or interference to the buried cable, or possibly to other cables in the immediate vicinity.

Moreover, visual identification of cables is not certain. Cables belonging to different utilities can be physically identical, so that visual location systems involve the need for some pre-knowledge of the cables in the vicinity of the excavation site.

It is also known to identify buried cables by applying an audio frequency electrical signal to the cable, and then detecting the magnetic fields generated by that signal at the surface. This removes the need for excavation, since the frequencies of the signal are chosen so that the magnetic fields generated will be detectable at surface level.

However, such an audio frequency signal may be transferred to other cables by induction and capacitive leakage. This leads to distorted magnetic fields, resulting in mis-location of the cable, or even the possibility of the wrong cable being identified, because induction results in the signal being carried by a cable other than that to which it is applied.

Many different proposals have been made for improving such electromagnetic location, but none have wholly prevented the problem of misidentification.

In our UK patent application number 9414847.5 we proposed that a very low, or zero, frequency current signal be applied to the cable, with the magnetic field generated by that current signal being detected by a suitable detector. The lower frequencies reduced the risk of the signal being transferred to other cables by induction and capacitive leakage. Frequencies less than 10 Hz, preferably less than 1 Hz were used. Although it was possible to operate at zero frequency (DC), there would then be the problem of distinguishing the signal from stray ground currents, and hence a frequency as close as possible to zero was proposed.

However, it has been realised that the system described in our UK patent application number 9414847.5 was not satisfactory when the buried cable branched. Since the branching would result in current division, the operator would not know the correct current level at any particular point on the branched cable, without knowing the branching on both sides of that point, so that correct cable identification may not be possible in a branched cable network using the techniques disclosed in our UK patent application number 9414847.5.

The present invention therefore proposes that signals are generated at ends of a branched cable network, with each signal having two components.

The first component is a current component for which the current has a known magnitude. The second component is a component identifying the end point from which the first component originated.

Such signals may then be repeated at periodic intervals, and the detection of those signals will enable an underground conductor to be identified on the basis of the signals, irrespective of the location of the detection point in a network of interconnected conductors.

To appreciate this, consider detection of the current levels at a random point in the network.

The signal detected at that point will be the sum of the signals applied from the ends of the network to which that point is connected. The second component of each signal will permit the detection apparatus to determine which ends are connected to the detection point. From the knowledge of those ends, and from the knowledge of the respective magnitudes of the current components from which end, it is possible for the detection apparatus to determine whether the sums of the first components correspond to the current magnitude predicted on the basis of the second components transmitted from each end to which the detection point is connected. If these agree, or are sufficiently close, then satisfactory identification can be made.If the measured magnitude differs from the predicted magnitude and more than a predetermined amount, then it can be said that the correct cable has not been identified unambiguously.

It should be noted that the present invention is particularly applicable to the identification of buried cables, such as telecommunication cables. In telecommunications, the conductor may be the sheath of a fibre-optic cable. However, the invention is not limited thereto, and can be used for the identification of any buried conductor, such as an underground metallic pipe.

Moreover, although the present invention was primarily devised for use with the low-frequency current signals of our UK patent application number 9414847.5, the present invention is not limited thereto. The present invention may therefore be used in combination with the known audio frequency identification systems. In either case, the signals generated by use of the present invention are superimposed on the other signals appearing on the conductor.

Preferably, the magnitude of the first component from each end of the network is the same.

Then, by knowing the number of end points connected to the detection point, it is easy to determine whether the current detected at the detection point agrees with that predicated, enabling accurate identification to be achieved. However, in some cases, it is not possible for each end point to transmit the same current, for example because the total current drain on the network would then be too great. In this case one or more end points may decrease the level of current that they transmit.

Then, the second component from that end point needs to contain information indicating that the current level has decreased. One simple way of achieving this is for the second component from each end point to be at a time within the overall signal which is different from the second component from any other end point. Then, by looking that the magnitude and location of the second component within each time interval, the detection arrangement can determine the magnitude of the first component which has been imposed on the network by the corresponding end point.

As has been mentioned above, the signals from each end point are repeated periodically. It is important that the signals from each end point are synchronised. The synchronisation may be achieved by transmitting a signal on the network which can be detected at each end point, and used for synchronisation. To prevent confusion with the current signals, that synchronisation signal is preferably a voltage signal. Moreover, although it is possible for the first component from each end point to be in the form of a single pulse, it is preferable for the first component to have a zero DC constraint, so that cross-correlation with any serious DC components of the received signal (e.g.

from the earth's magnetic field) will result in zero.

It is also preferable for the first component to have a sharp auto correlation function so that a receiver may discern it from noise in the general case where the receiver is not synchronized to the transmitted signal. Synchronization may then be achieved by cross correlating the received signal with a replica of the first component of the transmitted signal.

If such signals are used in combination with the low-frequency techniques discussed in our UK patent application 9414847.5, components of the signals must themselves have a low frequency. This way, the current amplitude is sustained throughout the length of the network, to a high percentage of its initial value. This will then mean that the components, and thus each signal, must have a relatively long duration but this is not a significant problem for cable identification, given that the cable has already been located (i.e. its position is known).

Embodiments of the present invention will now be described in detail, by way of example, with reference to the accompanying drawings, in which: Fig. 1 is a schematic block diagram of a cable network to which the present invention may be applied; Fig. 2 illustrates a signal generated by an end unit of the network of Fig. 1; Fig. 3 illustrates voltage pulses and signals appearing on the network of Fig. 1; Fig. 4 shows an end unit of the network of Fig.

1; Fig. 5 shows in more detail part of the end unit of Fig. 4; Fig. 6 shows basic principles of cable identification according to our UK patent application number 9414847.5; and Fig. 7 shows in more detail the way that cable identification is carried out in our UK patent application number 9414847.5.

Fig. 1 shows a cable network 10 with two branches 11,12. One end of the network is connected to a transmitter unit 13, and each of the other ends of the network 10 are connected to respective end units 14,15,16.

Each end unit 14,15,16 imposes a series of current signals on the network 10, to enable any cable forming a part of the network 10 to be identified.

Fig. 2 shows an example of the signal imposed on the network 10 by one of the end units 14,15,16.

The signal has a first component 20 and a second component 21. The first component 20 represents a series of pulses of known current magnitude and direction, and is the same for each of the end units 14,15,16. Thus, at any point in the network, the pulses of the first component 20 from all end points connected to that point in the network will add together, to form pulses which are the sum of all the end units 14,15,16 connected to that particular point in the network. It should be noted that the pulse sequence illustrated in the first component 20 of Fig. 2 is one chosen to provide a zero DC constraint, so that cross-correlation with any series DC component of the received signal (e.g.

from the earth's magnetic field) will result in zero.

The second component 21 comprises two pulses 22,23, each of which is of the same magnitude as the pulses of the first component. The pulse 23 is an end pulse, identifying the end point of the signal, and pulse 21 is an end-unit identification pulse.

The location of that identification pulse 21 relative to the end pulse 23 identifies which of the end units 14,15,16 generated the signal. Thus, as illustrated in Fig. 2, the pulse 22 is at position 23, and thus may correspond e.g. to the end unit 16.

A pulse at position 22 could then correspond to a signal from end unit 15 and a pulse at position 21 to end unit 14. A further pulse position 24 may correspond to a further end unit, not shown in Fig.

1.

Now consider a current detector located at or proximate position A in the network 10. That detector may, for example, operate on the basis of detection of magnetic fields generated by the current in the cable at that point A in the network 10, using e.g. a known electromagnetic locator.

Then, the current detected at point A will be the sum of the signals from end units 15 and 16. Since the first components 20 of each signal are summed, the current magnitude of the pulses during the first component 20 will be double that from any one end unit. The summing of the second component 21 will result in pulses at positions 22 and 23, so that the detector can determine that there are two end units 15,16 connected to the point A. Hence, the detector can predict that the magnitude of the current pulses during the first component 20 should be double that of any one end unit. If the magnitude of the pulses actually detected is such a double value, or within a predetermined range of that double value, then the operator of the detector can be confident that the cable has been correctly identified as one belonging to the network 10.If a different current magnitude is detected from that predicted, the cable identification is not reliable.

Similarly, at point B, the signals will be the sums of the current signals from each end unit 14,15,16 and there will be pulses at positions 21,22 and 23 of the end components 21. The predicted current magnitudes for the sums of the first component 20 will therefore be three times that from one end unit. Again, if the currents measured during the first component 20 are three times that from any one end unit 14,15,16, or within a predetermined limit of that, then again the cable can be correctly identified.

It may not always be possible (for example, due to voltage safety limits and/or cable resistance) for each end unit 14,15,16 to apply the same current to the network 10. If this is the case, one or more end units 14,15,16 recognises that it cannot transmit the standard current, and may then transmit a current of lower magnitude. It is then necessary to indicate to any detector that the current value has been reduced. For this reason, the second component 21 contains additional pulse positions 25 to 28. If e.g. the end unit 14 must transmit a reduced current, it does not signal at pulse position 21, but instead signals at pulse position 25. Hence, the detector can determine that one or more end units sent to the detector point are transmitting reduced currents.Typical actual values of current magnitude are 50mA in normal operation, with a reduced value of 20mA. The excitation voltage is preferably negative.

In order for such an arrangement to operate satisfactorily, the signals from each of the end units 14,15,16 must be synchronised. This is achieved by the transmitter 13 imposing voltage pulses on the network 10, as illustrated in Fig. 3.

The transmitter unit 13 normally imposes a voltage of -Va on the network 10, but this is reduced to zero periodically. That periodic reduction triggers the start of the signal from each end unit 14,15,16.

The lower part of Fig. 3 illustrates the relationship of these voltage changes with the components 20,21 of the signals from the end units 14,15,16.

As illustrated in Fig. 3, the duration between voltage pulses is 8.0s in this embodiment, giving a pulse width of approximately 0.25s. This is sufficiently long for the pulses to be effectively equivalent to DC currents on the network 10, so that attenuation is not a problem.

Moreover, for satisfactory operation, the transmitter unit 13 should also provide system integrity checks. This can be done readily by the transmitter unit 13 monitoring the pulses received from the end units 14,15,16. Those pulses are synchronised with the transmitter's output and, since they are of low frequency, can be separated by low-pass filtering from the signals on the network 10. DC earth leakage may be determined by making a measurement when the end units 14,15,16 are not applying pulses to the network 10. For example, in the signal sequence illustrated in Fig. 2, tests may be carried out during pulse positions 13 to 19.

Fig. 4 illustrates in more detail the structure of end unit 14. The structure of end units 15 and 16 is the same. In fact, the structure illustrated in Fig. 4 is conventional for the termination of a telecommunications cable, with the exception of the current unit 40. Thus, as illustrated in Fig. 4, the cable 41 from the network 10 is connected to ground via a parallel connection from the current unit 40, a first filter 42, a second filter 43, a lightening protection gas tube 44 and a switch 45.

The switch 45 is controlled by a control unit 46, which connects the cable 41 directly to ground in the absence of any signal being applied thereto.

The filter 42 passes signals to ground, and the filter 43 is a mains filter.

The structure of the current unit 40 is shown in more detail in Fig. 5. The current unit 40 has a current generator 50 controlled by a processor 51 and connected in series with a resistor 52. A detector circuit 53 detects the voltage pulses generated by the transmitter unit 13, and passes signals to the processor 51 which triggers the current source 50 to generate current pulses as illustrated in Fig. 2.

Initially, those pulses will be generated at the desired magnitude which is the same for each of the end units 14,15,16. However, if the processor 51 detects that the current actually generated is less than that value, by measuring the voltage appearing across resistor 52, then the current source 50 is controlled to produce the predetermined lower current value, as previous described. The processor also then controls the current source 50 to generate pulses at positions 21 or 25, depending on whether the current source 50 is transmitting the normal, or reduced current.

In order to power the current transmission, power is stored during times when a voltage is applied to the network 10 from the transmitter unit 13, and may be stored during any off-state.

As was previously mentioned, the present invention is particularly, but not exclusively, concerned with location of cables based on the lowfrequency signal described in our UK patent application number 9414847.5. The signals from the end units 14,15,16 are then superimposed on those low frequency currents.

For completeness, the arrangement disclosed in our UK patent application number 9414847.5 will now be described in more detail, with reference to Figs.

6 and 7.

Referring first to Fig. 6, a cable 110, which is buried underground, is to be identified. The cable 10 as illustrated on Fig. 6, is not branched and may correspond to two end points of the network of Fig. 1. Since there may be many other cables in the vicinity of cable 110 at particular sites, it is necessary not only to determine the position of the cable 110 (i.e. to "locate" the cable), but also to identify that the cable is that belonging to a particular utility (to "identify" it).

According to a this arrangement a signal generator 112 is contacted to the cable 110, as is a switch 114. Suppose now that the switch 114 is closed, permitting a current to flow along the cable 110. Since the same, or similar, generator 112 can be used, the frequencies of that current are again low, being less than 10 Hz. Again, frequencies not greater than 1 Hz are preferred, with the aim of getting as close as possible to zero, but permitting the current to be distinguished from stray DC ground currents.

Because of that low frequency, there will be little or no induction of the signal on the cable 110 to adjacent cables. However, the penetration of the magnetic field around the cable 110 will be small. As a result, if the probe 116 carries a magnetometer 130, the electromagnetic field around the cable 110 will be detected only when the magnetometer 130 of the probe 116 is sufficiently close to the cable 110. Distances of the order of 15 cm are preferred.

The electromagnetic field from the cable 110 can be detected by the magnetometer 130 independent of the material between the cable 110 and the magnetometer 130. Thus, in the arrangement of Fig.

7, the cable 110 can be identified by the magnetometer 130 even if there is soil or water in the space 124. On the other hand, the magnetometer 130 will detect any magnetic field of sufficient strength. Thus, if the cable 126 in Fig. 7 also carries a corresponding signal, this will also be detected by the magnetometer 130. There is thus very little chance of total failure of identification, but there is the possibility of misidentification.

In fact, the possibility of mis-identification is low because of the low frequency of the current signal. Since the frequency of that current signal is low, any inductive or capacitive coupling of the signal to adjacent cables is minimal. As a result, there is a substantially constant current along the length of the cable 110, and any stray fields from other cables are likely to be of lower amplitude, and hence less likely to be detected.

Because the frequencies of the current and voltage signals applied to the cable 110 by the generator 112 are low, a relatively long sampling period (e.g. of the order of 10s) is needed.

This invention is particularly, but not exclusively, useful in combination with the locator disclosed in our UK patent application number 9409003.2. In that application, a locator for locating an underground cable had a ground penetration probe containing locator antennas which detected electromagnetic signals from a conductor such as a cable, and enabled the position of the ground penetration probe relative to the conductor to be determined. The probe could therefore be driven into the ground towards the cable 110, and the operator provided with information that enabled the probe to brought into close proximity to the cable without the risk of the probe damaging the cable 110 due to forceful impact.

Using such a device, and with the sensor 118 and magnetometer 130 mounted on the ground penetration probe, it is possible to bring the probe into contact with the cable to enable measurements to be made.

The detection of the signals generated by the end units 14,15,16 may be carried out by the same magnetometer 130 as described above, as an additional cable identification step.

Claims (16)

1. A method of identifying a branched conductor comprising: applying signals to a plurality of points on the conductor, each signal having a first current component with a current of a known magnitude and a second component which identifies the corresponding one of the plurality of points at which the signal is applied; and detecting the plurality of signals at or adjacent a further point on the conductor, thereby to identify the conductor at that further point.

2. A method according to claim 1, wherein the signals are applied at periodic intervals.

3. A method according to claim 1 or claim 2, wherein the second component identifies the corresponding one of the plurality of points by the timing of a pulse of said second component.

4. A method according to claim 1 or claim 2, where the plurality of points are end points of the branched conductor.

5. A method according to any one of the preceding claims wherein the current of the first components of all of the signals has the same magnitude.

6. A method according to any one of claims 1 to 4, wherein the second component contains information indicating the magnitude of the current of the first component for at least one of the signals.

7. A method according to any one of the preceding claims, wherein the applying of the signals to the conductor is synchronised by a voltage pulse.

8. A method according to any one of the preceding claims, wherein the conductor is buried underground.

9. A method according to any one of the preceding claims, wherein the plurality of signals are detected adjacent, but spaced from said conductor, on the basis of electromagnetic fields generated by said signals.

10. A branched circuit network comprising: a branched conductor means for applying signals to a plurality of points on the conductor, each signal having a first current component with a current of a known magnitude and a second component which identifies the corresponding one of the plurality of points at which the signal is applied; and means for detecting the plurality of signals at or adjacent a further point on the conductor, thereby to identify the conductor at that further point.

11. A network according to claim 10, where the plurality of points are end points of the branched conductor.

12. A network according to claim 10 or claim 11, also including means for applying a voltage pulse to said conductor, and said means for applying signals to said conductor are arranged to detect said voltage pulse and to apply said signals in response to said voltage pulse, thereby to synchronise the application of said signals.

13. A network according to any one of claims 10 to 12, wherein the conductor is buried underground.

14. A network according to any one of claims 10 to 13, wherein said means for detecting the plurality of signals is arranged to detect electromagnetic signals generated by said plurality of signals adjacent but spaced from said conductor.

15. A method of identifying a conductor substantially as any one described herein with reference to Figs. 1 to 5 of the accompanying drawings.

16. A conductor network substantially as described herein with reference to and as illustrated in Figs. 1 to 5 of the accompanying drawings.