Fibre optic sensor method and apparatus

DESCRIPTION

Method and apparatus of a fiber optic sensor

Field of the Invention

[0001] This invention relates to a method for identifying an alteration of interest using a fiber optic interferometer and finding the location of the alteration using a separate sensor system. Alterations of interest may be, for example, the breaking of reinforcement cables in a concrete pipe, the rupture of the cables within the suspension cables, a pipe leak, or an intrusion by a human or a vehicle.

Background of the invention

[0002] There are fiber optic sensors that can control events at a distance of twenty kilometers or more and that can operate with relatively low power. Such fiber optic sensors can detect 10 acoustic and seismic alterations, such as intruder footsteps near a controlled perimeter, noise associated with intentional damage to a controlled piece of infrastructure such as an electrical or communications cable, noise of a leak in a pipe, or the breaking of a reinforcement cable in a concrete pipe or a cable in a bridge cable. Some of these sensors have detection grids spaced by protected parts, so that the location of an alteration can be found by determining in which grid (s) the alteration is detected. Others use a pulsed laser light, where the reflected signals caused by an alteration are reflected back to the origin and the location from which the signals come is determined by the time span from the impulse to the reception of the reflected signal.

[0003] Such fiber optic sensors have not been very effective, since many different types of alterations can trigger a response. Once a response is triggered, the location from which it comes 20 must be investigated to determine if a condition that requires corrective action is present. In addition, sensors that depend on the reflection of an impulse may lose or misunderstand transient effects that have their maximum effect at a time when the impulse is not scanning the particular location when they occur.

[0004] Interferometry sensors are known to be sensitive to measuring for a long length, for example, the total length of the fiber optic sensor. Because the total length or a long length in the area of ​​interest is sensitive to the measurand, a signal indicating an alteration is acquired at or very close to the source of the alteration. This provides an advantage in the signal-to-noise ratio, since the sensor does not move longitudinally from the alteration along the fiber, as is the case when there are fixed spaced sensors. Because the distance from the detection point closest to the source of the alteration is minimized, the deterioration of the signal-to-noise ratio relating the signal attenuation to the distance is also minimized. 30

[0005] Interferometry sensors are known in the art and several types are known, such as a Sagnac effect interferometry sensor shown in Udd USP 5,636,021 or a Michelson interferometry sensor as shown in Jones et al USP 4,725. 143.

[0006] Finding the location where the alteration occurred along with the length of an interferometry sensor is difficult. Udd (USP 5,636,021), Tapanes et al. (USP 6,621,947) and Kyoo, Juarez and Taylor ((2003) SPIE, Vol. 5090. Pp 131-35 141 have attempted to achieve the location of the alteration using radii of beam amplitude against propagation (Udd), arrival times of alterations in circuit interferometers (Tapanes) or phase-sensitive or optical domain reflectometry temporary (Kyoo, Jurarez and Taylor). However, the proposed ways of finding the location did not work well. In Udd, if the return circuit of the Sagnac circuit is affected by the alteration, the radiometric approach used to estimate the location does not work well and it is usually not possible to know if the return circuit is affected. In Tapanes, the inability to follow rapid variations (slew rate) of the signals resulting from an alteration makes the location of the source difficult.

[0007] In the case of methods based on time domain reflectometry, such as that of Kyoo, Fernades and Taylor, the location of the alteration can be determined by looking at the point along the return signal of an impulse where it is altered , or where the disturbance begins, resulting from the alteration. However, the use of such a pulsed laser 45 means that there is no continuous control. Instead, each location along the fiber optic cable is only controlled at times when an impulse passes through it. In addition, especially in long sensors, there is considerable noise and only limited bandwidth is available. Continuous control can be very important when one is detecting an evanescent event or an event where the "signature" measure changes rapidly over time, making it impossible to deduce what caused the event without full registration. In addition, the reduced bandwidth usually provides insufficient information to characterize the received signal, in order to assess its possible cause.

[0008] Some examples of evanescent events include:

- the acoustic signal caused by the rupture of a reinforcement cable in a concrete water pipe wrapped with reinforcement cables.

- the acoustic signal caused when a reinforcement cable in the bridge cable is split.

- the landing of an object, such as a thrown object, which has been pulled towards a perimeter controlled by a perimeter intrusion detection system. 5

[0009] The known fiber optic sensors do not identify the location from which a signal comes and provide at the same time sufficient information to make a reliable identification of what causes the signals. In this sense, they are inferior to existing systems that are not fiber optic. For example, Paulson USP 5,798,457 uses acoustic or seismic detectors in a set to detect signals and analyze both the location from which the signal comes and its characteristics to see if it is indicative of a condition such as a cable break. 10

[0010] Crawford (US 5355208) publishes a fiber optic detector circuit to detect alterations comprising a closed circuit in which two anti-propagation beams are forced to circulate in opposite directions. The two Crawford beams are distributed interferometry sensors capable of detecting alterations along their total length at the same time. Crawford does not describe or claim a second sensor with the properties of detecting both alterations and their locations. Instead, Crawford uses the temporal relationship between the arrival of the main edge of an alteration in each beam against propagation to determine the location of an alteration, which has important limitations.

Brief Description of the Invention

[0011] The invention uses two separate sensors.

[0012] The first is a fiber optic interferometer, which detects all parts of the length to be detected and does so continuously. This can detect evanescent effects. In addition, it detects alterations that produce signals anywhere within a wide bandwidth. The information can be extracted to indicate the frequency distribution by the time the signals are controlled, thus providing a good indication of what caused the signals.

[0013] Suitable fiber optic interferometers are, for example, Sagnac effect interferometers and Michelson interferometers.

[0014] The output of signals from the fiber optic interferometer is examined, for example by computer, to see if they are apparently representative of an "event of interest". An event of interest is a signal or group of signals with one or more predetermined characteristics that are likely to have occurred as a result of a condition for which a check is required. 30

[0015] An "event of interest" is predefined by the system operator, for the particular sensor system. Generally, the operator will want to investigate any signal or group of signals that has more than one particular energy peak or more than one energy peak in a chosen frequency band or bands. The power peak can be displayed on a screen using an analog or digital signal. The minimum energy peak that is of interest can be determined by testing events on the current system or a test system or other system installed with similar characteristics, observing the signal output of said events and choosing a lower energy peak than the output through these events as an energy limit. Any signal that shows an event with an energy greater than the limit is then defined as an event of interest.

[0016] Test events can be chosen with respect to the type of risk being controlled. For example, in a system designed to detect cable breaks in prestressed reinforcing cables for concrete pipes 40 wrapped in prestressed cables or reinforcement cables for a concrete floor in the construction, some reinforcement cables can be deliberately cut to observe the response . Where the system is designed to detect damage on a bridge, several test operations can be carried out, such as a simulated vandalism attack on a post or cable, the breaking of a cable on a bridge cable and a vehicle hitting the bridge . In each case, the answer is observed. When the system is designed to detect intrusion, the response is noted for the passage of a human and the passage of a vehicle. From the tests, a peak energy level is chosen low enough to include all probable events that are considered of interest by the system and any event that shows a signal with as much energy or more is defined as an event of interest . If desired, the data can also be collected by the characteristic shapes of the signal graph and how quickly the signal decays at particular events. The spectral densities of the signals at different wavelengths of different types of 50 events can also be determined. This provides data that can be used to display signals initially chosen as events of interest with a view to exclude some that, from the extra data characteristics, seem to be explainable because they were caused due to an event that is not interest in the context of control.

[0017] The second sensor is a location sensor. This produces signals from substantially the same sensor length as the interferometer. In normal operation, the signals received by the location sensor are maintained for a suitable period of time and then discarded. However, when the signal output is determined by the interferometry sensor as an event of interest, then the signals from the location sensor during the period in which the event of interest occurred and preferably for a period of time before 5 and After the event of interest they are saved and examined to see if there is an abnormal activity at any location at that time or nearby. Once the location is known, the operator (or the automatic system) can reevaluate if the event is an event of interest or not, based on the location. For example, a sound characteristic of a passing truck could be considered to be not an event of interest if it comes from a location along a sensor system that detects intrusion near the road, but the same signal could be considered an event of 10 interest when it comes from a sensor location where trucks are not expected to be.

[0018] Thus, the outputs of the two sensors are related to each other so that the user can associate an alteration position as calculated by the location detection sensor with a particular set of characteristics determined by the interferometry sensor.

[0019] The location sensor is also a fiber optic sensor but one that gives the location of alterations that it detects, such as an OTDR phase sensor or a Brillouin effect sensor. The location sensor may be on the same optical fiber as the interferometer, or it may be on a separate optical fiber adjacent to the fiber optic interferometer. Preferably, it will be in a separate optical fiber located on the same cable as the interferometer sensor.

[0020] In accordance with the invention, the identification of an event of interest is discovered by the interferometry sensor of greater bandwidth. If the sensor is installed in an environment where too much noise is generated, then recognition of a particular event that contains features of interest is needed. When such an event is recognized, then the position of the event source must be distinguished. This is done by examining the temporal appearance and extinction of some characteristics of the event on both sensors. Since each event will affect both sensors during the same time, they start at the same time and end at the same time if the sensors are in the same position relative to the structure they are controlling, then a temporal correlation of the appearance and disappearance of the characteristics that distinguish said event from background noise should allow the event as it appears on each separate sensor to be identified as an event of interest.

[0021] For example, in a prestressed water pipe according to Paulson debate (USP 6,082,193), a distributed interferometry sensor and a phase OTDR can be constructed in an optical fiber installed in the pipe. Current noise, traffic disturbances and other features will produce noise effects on both sensor tracks. The failure of a prestressed cable would generate a burst of sound that would travel through the water until the detection fiber was found, altering it. The acoustic wave would also travel in both directions along the pipe, slowly attenuating and resulting in a slow disappearance of the alteration in both sensing pathways.

[0022] By continuously monitoring the interferometry sensor, the event can be recognized as one of interest. An information buffer indicating the location sensor output (eg, the phase OTDR sensor) is maintained. 35 When an event of interest is determined from the interferometry sensor, reference is made to the results of the phase OTDR of the moments just before and during that occurrence of the anomalies of interest in the interferometry sensor. The outputs of the phase OTDR are then used to determine the location along the sensor of the anomaly of interest.

The drawings 40

[0023] The invention will be described in more detail with respect to the following drawings, in which:

Figure 1 shows a first embodiment of the invention in which both sensors are located in the same fiber of the optical fiber.

Figure 2 shows a second embodiment of the invention in which the interferometry sensor and the location determining sensor are located in separate but adjacent optical fibers. Four. Five

Figure 3 shows the output of an interferometry sensor recorded over a period of time indicated in milliseconds, containing a detected event.

Figure 4 shows a Fourier transformation of the interferometry sensor output shown in Figure 3.

Figure 5 shows the signal output of the location sensor for a period of time just before, during 50 and after the event recorded in Figure 3.

Detailed description of the invention

[0024] The invention will now be described in more detail with respect to the drawings.

[0025] Figure 1 shows a first embodiment, in which the interferometry sensor and the location sensor are located on the same optical fiber. The equipment related to the generation of the interferometry sensor beam and for the reading of signals received by the interferometry sensor is generally indicated in 1 and 5 said equipment has numerical references between 100 and 199. The equipment related to the location sensor is generally indicated in 2 and said equipment has numerical references between 200 and 299. The common equipment for both and the detection part is generally indicated in 3. Other features present, such as the detected environment, are indicate with numerical references greater than 400.

[0026] The part of the apparatus that forms the interferometry sensor beam will be described first. The exemplified interferometry sensor 10 is a Sagnac type interferometer. The particular Sagnac type interferometer used in the example here was constructed by Pure Technologies Ltd., 705 11th Ave. SW, Calgary AB, Canada. Sagnac type interferometers are marketed by Blue Road Research, 219 NE 219th Avenue, Greshem, Ore, USA

[0027] Alternatively, a Michelson interferometry sensor can be used. This is also marketed by Blue Road Research, whose address has been given previously.

[0028] In the exemplified interferometry sensor, a coherent light source (a laser) 100 and a suitable control circuitry 110 for this is provided. In a particular example, the laser is chosen to give coherent light at a wavelength of approximately 1310 nm. However, this wavelength can be selected in a known manner according to the particular type of fiber to be used and the expected sources of interference to minimize the likely interference and to maximize the signal of any expected disturbance that is sought to control it.

[0029] The laser signal is sent through a suitable optical fiber 120 to a connector 130. The coupler 130 divides the laser output into two parts, which pass through optical fibers 121 and 122. Fiber 122 leads to a modulator. 140, which modulates the laser light that passes through it. The modulated light is then sent through a delay coil 150 and back via line 123 to a coupler 160, the modulated light coming along the branch 123 and the unmodulated light from the fiber 121 come together of the coupler 160 in the fiber 124. They pass through a short pass filter 170, which has the function of filtering the wavelength longer than the wavelength of the laser 100. For example, in this example, where the laser 100 has a wavelength of 1310 nm, the filter filters the wavelengths longer than with 1310 nm, with the filtering becoming more effective as the wavelength is It is longer. The filter is used to eliminate strange wavelengths of the laser signals that return and are directed to the receiver. From the filter 170, the fiber 125 leads to a combination coupler.

[0030] The part of the device that forms the location sensor is now described. In the example shown, the apparatus is an OTDR phase sensor, as described in Taylor (SPIE, 2003). Any other laser sensor that is capable of detecting an alteration of the controlled type and its location could be used, such as a system based on 35 Brillouis available as a Ditest Model BOTR marketed by Smartec SA located on Via Probette 11, CH6928 Manno, Switzerland.

[0031] The exemplified location sensor has a laser 200, suitably controlled by the controls 210. In one example, the coherent laser light is at a wavelength greater than the laser light 100. In the example, the wavelength is 1550 nm. The light passes through the optical fiber 220 to an impeller 230, which 40 is controlled by suitable controls 235. If desired, the fiber 220 may contain an attenuator (unnumbered) known in the art. The impeller 230 causes the coherent laser light to be sent in pulses, at a desired repetition rate. The pulses are sufficiently spaced so that the light can travel to the end of the sensor and return before the next pulse is sent. Suitably, a pulse may be of the order of 20 to 1000 nanoseconds in length and the repetition (update) rate should be at least 10 times per second, depending on the length of the sensor. Preferred speeds are greater than this, between 500 and 5000 kH. The sampling rate of the return waveform will be chosen to obtain a spatial resolution of 0.2 km or less (that is, a seen event can be located within 200 meters). Preferably, an even smaller spatial resolution will be chosen. Pulsed lasers are well known in the art and an adequate pulse length, the repetition rate and the sampling rate for the particular length 50 and the composition of the sensor being used can be chosen by a person skilled in the art.

[0032] In water pipes, the propagation of the alteration in both directions from the source offers the opportunity to greatly improve the accuracy of the location estimate, using the expected symmetric propagation model over the many traces of the phase OTDR to more accurately measure the position of the

origin of the alteration and eliminate from consideration the parts of the trail that result from other noise in the pipe. Thus, in water pipes and similar applications, the spatial resolution that is not as good as 200 meters can be tolerated, since the symmetric phase OTDR model can be used to improve accuracy.

[0033] Pulsed light passes through a fiber 221 and preferably through an amplifier 240 outside it along the fiber 222. If the amplifier is not present, the fiber 222 is a mere continuation of the fiber 221 . 5

[0034] The light then passes to a circulator 250. This is known in the art of fiber optic components. Two different fibers, 223 and 224 also enter the circulator. Fiber 223 is directed towards receiver 260 and fiber 224 is directed to a long pass filter 270.

[0035] The circulator, as is known in the art, allows light to pass from fiber 222 to fiber 224, or from fiber 224 to fiber 223. 10

[0036] The long pass filter 270 filters the light having a shorter wavelength than the light of the coherent light source 200. Thus, with a light source 200 leaving at 1550 nm, the long pass filter it could filter the light with a wavelength shorter than 1550 nm., the filtering becoming more pronounced the shorter the wavelength becomes.

[0037] The long pass filter light passes along line 225 towards the combination coupler 300. 15

[0038] In the combination coupler 300, the light that has been generated with the laser 100, both unmodulated and modulated, is sent along the same fiber as light from the line 225. This fiber is indicated as 310.

[0039] All the optical fibers mentioned so far, fibers 120, 121, 122, 123, 124, 125, 220, 221, 222, 223, 224, 225 and 310 and 336 and 227 (to be described) are preferably protected from Adequate way to minimize ambient noise to reduce interference in sensor systems. The protection can be by physical separation, or by the use of materials that will block the alterations, as is known in the art.

[0040] Fiber 310 joins the fiber of detector 320. Fiber 320 is not protected, so it is capable of having alterations created in the interior light due to an external alteration.

[0041] Suitably, the fiber of the detector 320 can be quite long, for example up to 20 km. or more. To indicate that the total fiber length of the detector is not shown, two parallel lines 400 to 25 have been drawn through the sensor fiber. These do not indicate an interruption in the fiber, but simply the fiber extends over a long distance and a part of it has been omitted. Fiber 320 terminates in terminator 330. The terminator has two parts. One is a "mirror" that reflects the 1310 back light. The other minimizes the reflections of 1550 nm of light. These are the normal functions of the terminator for an interferometry sensor and a phase OTDR sensor respectively. 30

[0042] The light that passes from fibers 225 and 125 to fiber 310 and 320 continues through the length of the fiber towards terminator 330. Whatever the reflected light back passes through a coupler 300 towards the filters 270 and 170. Filter 270 excludes the light from the interferometry sensor, since that light has a wavelength of 1310 nm and cannot pass through filter 270. The filter 170 excludes the pulsed laser light since the light has a wavelength of approximately 1550 nm, which cannot pass through the filter 170. 35

[0043] Taking first the light that can pass through 170, it passes to the coupler 130, after which it proceeds along the fiber 126 to a receiver, analyzer and demodulator 180. The receiver-analyzer-demodulator also receives the modulation signal on fiber 127, so that you can access the interference of light beams against propagation. The receiver and the analyzer use the light interferometry model in a conventional way to create a digital or analog output indicative of the alterations that have occurred along the fiber. An operator console 190 allows an operator to view the signal and direct more operations on it, such as a Fourier transformation or other analysis.

[0044] Similarly, the light passing through the filter 270 is directed towards the receiver 260, which produces an output related to the amplitude of the light and the elapsed time, from the start of the pulse. This is conveniently extracted as a three-dimensional batch, with an elapsed time recorded from successive pulses, the delay time for OTDR reflections and amplitude.

[0045] In Figure 1, the sensor 320 is shown schematically resting on the bottom of a concrete pipe generally indicated as 410. The concrete pipe has a wall 420 containing the prestressed cables 430. Only about Few of those cables are shown in the drawing, but it is understood that those cables could be wrapped tightly around the pipe to keep it compressed. In the example, a cable 431 is broken at 432. As a result, acoustic and seismic waves (both generally shown as

433) expand outward from the break. The waves can be transmitted through the middle of the concrete pipe, or through the concrete pipe and through the fluid medium it contains. These waves first affect the sensor at a point 450. Subsequently, the waves also affect the sensor at locations that extend in both directions from point 450, as the rear parts of the acoustic and seismic wave fronts hit the sensor 320. 5

[0046] When the wavefront hits sensor 320 at 450, and subsequently hits it at locations that extend outward from location 450 in both directions, this causes disturbances in the light waves in the fiber optic fiber 320, both in the 1310 nm light of the interferometry sensor and in the 1550 nm light of the pulsed detection system. The signals generated by these disturbances travel along fiber 320 and along fibers 125 and 225. Filter 170 filters the 1550 nm wavelength and filter 270 filters the 1310 10 nm wavelength. Thus, each receiver 180 and 260 receives substantially only the light generated by its associated laser, thus making it easier to analyze the alteration of the light that reacts to an alteration in the fiber.

[0047] Figure 2 shows an alternative form of the system. Similar numbers are used for similar parts.

[0048] The difference between Figures 1 and 2 is that the interferometry detector laser and the position detector laser are not connected to the same sensor. On the contrary, there is a sensor 320a for the interferometry sensor and a sensor 320b for the laser position detector. Each also has a separate terminator, numbered 330a and 330b respectively. For mirror 330a a mirror is used. For terminator 330b, a non-reflective terminator known in the art is used. In addition, filters 170 and 270 and coupler 300 are not necessary since the light of the two lasers is not sent through the same fiber. In fact, it is no longer necessary to use different wavelengths of light. For example, in FIG. 2 the embodiment, both laser 100 and laser 200 20 could use 1550 nm light if desired.

[0049] The system of Figure 2 is preferred, because if there is a problem with one of the sensors, it can be removed and replaced without altering the other sensor (as long as the two sensors are not on the same cable). In addition, it allows you to choose an optimal wavelength for each sensor, without worrying about having to have a sufficient separation between the wavelengths to prevent signals from overlapping. 25

[0050] In the embodiment of Figure 2, the two sensors 320a and 320b must be adjacent to each other, so that each one receives a signal of an event of interest, such as a cable break or a noise made by an intruder at the same time substantially. This is necessary so that the location can be determined accurately using the location determination system once the event of interest has been determined, using the interferometry system. 30

[0051] The distance between the two sensors may vary according to the installation. However, it is generally preferred that the two optical fibers of the fibers 320a and 320b be adjacent to each other as for example within a single cable containing multiple fibers. In any event, it is preferable that there is no separation of more than 50 centimeters between the two fibers and preferably not more than 10 centimeters.

Example 35

[0052] The use of the apparatus to determine the events of interest will be described with respect to a test facility. The test facility was established as Figure 2. Sensor 320a had a sensor length of 10 kilometers and was placed on a disused concrete water pipe. The sensors 320b and the sensor 320a were two individual fibers within a single fiber optic cable and were the same length.

[0053] The interferometry sensor had a wavelength of 1310 nm. The pulsed OTDR sensor had a wavelength of 1550 nm and was set to scan the fiber optic length every 10 milliseconds. Each scan took 200 microseconds. At least 750 OTDR scans were retained in memory. As each new scavenger was added, the oldest was discarded.

[0054] For the purpose of the test, it was determined that an event of interest would be any event that gave an analog signal of more than 5 volts at an analog output in the particular test facility. The analog output 45 was an arbitrary representation of the alterations in the interferometry model in the Sagnac interferometry laser. An initial test showed that cutting a reinforcement cable in a concrete pipe gave a signal of 10 volts or more. Since the test was to see if the cable break could be detected, it was decided that the limit for an event of interest would be set at approximately half that level, or at 5 volts.

[0055] Figure 3 shows, as an analog signal output of volts relative to time, the output of the interferometry sensor as recovered in receiver 180. The deviations of the center line represent a phase change in the light of return. The time is arbitrary, from a base 0 that is common to both sensor outputs. It will be noted that the interferometer essentially provides a flat signal 600 (there is no phase change between the light

sent and light received), up to approximately 30 milliseconds on the arbitrary scale. Then, it provides a long answer 610. The response decays rapidly, but there are still more disturbances in the signal than there were before the event that occurred at 30 milliseconds. Approximately 126 milliseconds on the arbitrary scale, an additional event 620 that interrupts the light occurs.

[0056] The event that was triggered at approximately 30 milliseconds is greater than 5 volts on the arbitrary scale 5 and is therefore an event of interest. Thus, the location location laser output (OTDR output) for at least 30 milliseconds (150 sweeps) prior to the beginning of the apparent event of interest is saved, such as the at least 600 subsequent neighborhoods.

[space between paragraphs added]

[0057] If it is desired to be more certain that the event is in fact an event of interest, the operator may then decide to make a Fourier transformation of the signal from the interferometer from the moment just before, during and after the event. Said Fourier transformation is shown in Figure 4. This shows that there is a considerable increase at the time of the event in the amount of wavelength energy below 4 KHz. However, this is only indicative of an impact that has hit the sensor and is not very useful for characterization. However, there is also an increase in energy in the wavelengths between 4KHz and 15 8 KHz, which is typical of the model seen when the prestressed cable is broken in a concrete pipe. Thus, the Fourier transformation confirms that the event is indicative of a cable break, which is an event of interest when controlling a concrete pipe with cable wrap.

[0058] Because this is an event of interest, the location must be determined. Figure 5 is a graph of the output of the OTDR location sensor. The graphic is in three dimensions. The scale to the right shows the time 20 elapsed in milliseconds. The scale on the left shows the scan time of the OTDR. Each line through the graph parallel to the scale is the trace of a scan. It will be remembered that in this example the scans are repeated every 200 microseconds. The vertical axis (showing the peaks) is an arbitrary scale that shows deviations from ordinary background noise. Conveniently, this scale can be expressed in volts, but the magnitude is not of interest in the particular application, except to determine that the event has passed the arbitrary minimum to be an event of interest. Instead, only time on the arbitrary scale of milliseconds is of interest.

[0059] Since an event of interest has been detected in the interferometry scale at approximately 30 milliseconds (arbitrary scale), a dashed line has been drawn along the graph at this point (the line is indicated with 500). A series of peaks are discovered starting at approximately line 500, these peaks have a duration of approximately 10 milliseconds. A line 510 is drawn to project the principle of these peaks on the OTDR scale. It is discovered that the line crosses the OTDR scanner scale at approximately 87 microseconds in the scan.

[0060] In other words, the disturbance is noted when the ODTR pulse has passed below the fiber for 87 microseconds. It is known that the particular fiber transmits light of this wavelength at 9.73 microseconds / km. Thus, the location of the disturbance corresponds to a position of 87 / 9.73 km, which is 8.94 kilometers, from the beginning of the sensor 320a and 320b.

[0061] The location 8.94 km below the sensors, in the concrete water pipe, was examined and found to correspond to a location where an induced cable rupture had occurred during the test. 40

[0062] It will be noted that there are several comparable events in the OTDR, approximately 85 milliseconds and approximately 121 milliseconds. These have been indicated as 520 and 530. Since event 520 had no corresponding peak in the interferometry sensor, this could be considered noise. Event 530 had a corresponding small peak 620 at approximately the same time of recording (Figure 3) of the interferometry sensor output. This is below the limit of what has been defined as a significant event. Therefore, normally, it would not be considered an important event. However, in this case, because it is closely linked in time to an event of interest, it may be desirable to examine the signal 620 at the additional interferometry output. For example, a Fourier transformation of that signal could be carried out to see if a signature characteristic of an event of interest could be noticed. Alternatively, the event referenced by the numbers 530 and 620 may be considered not of sufficient interest (due to the relatively small interferometry output) for further consideration.

[0063] It will be noted that the event at 520 of the OTDR output does not have a corresponding event at the interferometry output. It often happens that there is a signal in one or another interferometry signal or OTDR, but not

there is a corresponding output in the other. In such cases, the signal is understood as a noise and is ignored, since it has not been confirmed by a signal on the other sensor.

[0064] Other additional embodiments of the invention will be apparent to those skilled in the art and it is intended that the appended claims be construed to include all such variants.