Techniques for enhancing detection and alerting for events in fluid pipes

Abstract

An apparatus (100) for monitoring a fluid pipe (1). The apparatus (100) comprises at least one sensing fibre (10) provided within the pipe (1), a light emitter (101) for introducing light pulses with particular characteristics into the fibre (10), a light detector module (110) configured to detect backscattering of the light pulses from the sensing fibre (10) in multiple different sensing modes and output a backscattering signal in response thereto and a processing unit (103) configured to process the backscattering signal to identify backscattering signal features characteristic of a particular event. Upon detecting an anomalous event or an inconclusive event (512, 514) the processing unit (103) is configured to control (518) the light detector module to switch between the sensing modes and / or to control (516) the light emitter to vary the characteristics of the emitted light pulses.

Description

[0001] Techniques for Enhancing Detection and Alerting for Events in Fluid Pipes

[0002] Technical Field of the Invention

[0003] The present invention relates to monitoring of fluid pipes. In particular, the present invention is directed to monitoring of fluid pipe integrity, the monitoring of fluid flow within a fluid pipe and / or leak and other in-pipe or near pipe event detection. The present invention further relates to an apparatus and a method for monitoring a fluid pipe as well as a network comprising one or more pipes monitored using the apparatus or the method of the present invention.

[0004] Background to the Invention

[0005] Many modem services rely upon a network of pipes to carry or distribute fluids. Examples include fresh water, waste water and sewage, industrial and mining, and fuels such as oil or gas. It is common to monitor the operation of the network and the condition of pipes. In this manner, blockages, leaks or other issues can be alerted, identified and scheduled for repair where appropriate.

[0006] Where pipes are provided above ground, monitoring may be achieved by visual inspection of the pipe exterior. In many cases, pipes are not accessible to visual inspection, being buried underground. Accordingly, pressure / audio sensors or the like may be utilised to detect vibrations of the pipe and thereby provide information on conditions within a pipe.

[0007] In particular implementations, distributed acoustic sensing (DAS), otherwise referred to as distributed vibration sensing - DVS has been used for monitoring pipes. DAS involves the detection of backscattering of light pulses introduced into an optical fibre. The time of arrival and intensity of the backscattered light is measured for each pulse, the time at which the backscattered light is detected being related to the distance along the fibre the light has travelled before being scattered. Subsequent changes in the reflected intensity of successive pulses from a common region of the fibre correspond to variations in the strain applied to the fibre at that region, for instance due to vibrations experienced by the region of fibre. In this manner, the DAS fibre can act as a plurality of virtual microphones along the length of the fibre and can locate events causing acoustic signals down to an accuracy of around 1 meter. One example of this technique is our prior application WO2019 / 166809.

[0008] Nevertheless, there are some factors that can limit the reliability or sensitivity of leak detection using a DAS mode. In particular, relatively small leaks or events may generate relatively small effects and thus are difficult to detect. Additionally, external noise and vibration can be transmitted through the pipe, masking any leak signals. This might be particularly common in urban areas where there are multiple other sources of mechanical vibration including traffic, industry and the like. Accordingly, this reduces the confidence level in detection of leaks or events using a DAS mode and / or the speed of accurate determination that a leak has started.

[0009] Additionally, by only using the DAS mode to monitor fluid pipes, some anomalous events may be misinterpreted. In one example, opening of a valve of the pipe may be misinterpreted as a leak.

[0010] It is an object of the present invention to provide methods and apparatus which at least partially overcomes or alleviates at least some of the above problems.

[0011] Summary of the Invention

[0012] According to a first aspect of the present invention, there is provided an apparatus for monitoring a fluid pipe. The apparatus may comprise at least one sensing fibre provided within the pipe. The apparatus may comprise a light emitter for introducing light pulses with particular characteristics into the fibre. The apparatus may comprise a light detector module configured to detect backscattering of the light pulses from the sensing fibre in multiple different sensing modes. The light detector may be configured to output a backscattering signal in response to the backscattering of the light pulses from the sensing fibre. The apparatus may comprise a processing unit. The processing unit may be configured to process the backscattering signal to identify backscattering signal features characteristic of a particular event. The processing unit may be configured to, upon detecting an anomalous event or an inconclusive event, control the light detector module to switch between the sensing modes. Additionally or alternatively, the processing unit may be configured to, upon detecting an anomalous event or an inconclusive event, control the light emitter to vary the characteristics of the emitted light pulses.

[0013] By providing a sensing fibre within a pipe, good acoustic coupling between the sensing fibre and the pipe (and any fluid within the pipe) is assured and the sensing fibre can be used to detect pipe condition information including pressure waves, negative pressure waves, flow noise, temperature changes, strain changes within a body of the pipe, orifice noise or the like.

[0014] Switching the detection modes and / or varying the characteristics of the emitted light pulses in the apparatus for monitoring the fluid pipe offers significant advantages in increasing the confidence level of detecting and classifying particular events, such as leaks. By switching between multiple sensing modes (upon detecting an anomalous event or an inconclusive event), the system can gather diverse types of data, each providing unique insights into the physical state of the pipe. This multimodal approach allows for cross-verification of detected anomalies or inconclusive events, enhancing the accuracy and reliability of event identification. When an anomalous or inconclusive event is detected, adjusting the characteristics of the light pulses can refine the resolution and sensitivity of the measurements. This dynamic adaptation ensures that subtle or complex signals are better captured and analysed, reducing false positives and negatives. Overall, these strategies enable the system to more confidently and accurately classify events, leading to more effective monitoring and maintenance of the fluid pipe infrastructure.

[0015] This adaptive approach is particularly advantageous because it optimizes the use of resources by not requiring the constant use of all sensing modes or continuous adjustment of light pulse characteristics. For events that are unambiguously detected and classified with high confidence in a single mode, there is no need to switch modes or modify the light pulses, thereby conserving energy and computational resources. This targeted response ensures that the system remains efficient and focused, applying its advanced capabilities only when necessary to resolve ambiguities or confirm uncertain events. The fluid pipe may comprise a pipe carrying any suitable fluid. In particular, suitable fluids might include but are not limited to: water, waste water, sewage, and fuel such as oil, gas, distillates or the like and chemical or mining products.

[0016] The light emitter may be a laser. The emitted light may any suitable wavelength for transmission along and backscattering within the sensing fibre. The light emitter and light detector may be integrated into a light transceiver unit. Suitable wavelengths may include visible, infrared or ultraviolet wavelengths.

[0017] In some embodiments, the light emitter, light detector module and processing unit may be integrated into a pipe sensor unit. Such a pipe sensor unit may be provided with a user interface. The user interface may enable a user to control operation of the pipe sensor unit and / or review indications relating to the condition of the monitored pipe.

[0018] The particular event may comprise one or more of: a leak in the pipe, a change in external conditions proximal to the pipe, third party intrusion events, the pipe bursting, a structural deformation of the pipe, a change in fluid flow within the pipe, a change in fluid pressure within the pipe, a blockage in the pipe, and / or valve and pump operation.

[0019] The change in external conditions proximal to the pipe may comprise ground movement or soil displacement near the pipe. Such movements may be indicative of landslides, construction activity, or natural settling of the ground, all of which may impact the integrity of the pipe. The change in external conditions proximal to the pipe may comprise detection of temperature changes in the surrounding environment, which could indicate seasonal variations, underground fires, or proximity to other heatemitting infrastructure. The change in external conditions proximal to the pipe may comprise vibrations caused by nearby machinery or heavy vehicle traffic. Advantageously, detecting these changes in external conditions allows for proactive maintenance and risk mitigation, ensuring the continued safe operation of the fluid pipe. The change in external conditions proximal to the pipe may comprise drilling or manhole lifting near the pipe. Third party intrusion events may comprise unauthorised access or tampering with the pipe by unauthorised persons.

[0020] The valve operation may comprise opening of a valve to allow fluid to flow through a previously closed section of the pipe. The valve operation may comprise closing of a valve to stop fluid flow. The valve operation may comprise partial valve adjustments, such as throttling to control the flow rate. Monitoring these valve operations ensures efficient and safe management of fluid distribution within the pipe system.

[0021] The structural deformation of the pipe may comprise bending of the pipe due to external pressure or ground movement. The structural deformation of the pipe may comprise corrosion or erosion over time that can thin the pipe walls. The structural deformation of the pipe may comprise buckling due to thermal expansion or contraction. Monitoring these structural deformations is crucial for early detection of potential failures, ensuring timely maintenance and prevention of more serious damage.

[0022] The processing unit may be configured to process the backscattering signal to distinguish between different types of leaks in the pipe. The different types of leaks in the pipe may comprise one or more of: pinhole leaks, longitudinal cracks, and circumferential breaks.

[0023] The ability of the processing unit to distinguish between different types of leaks provides several significant advantages for pipe monitoring and maintenance. By dynamically adapting the sensing strategy (e.g., by dynamically switching between the sensing modes and / or varying the characteristics of the emitted light pulses), the system can ensure high accuracy and reliability in leak identification. By accurately identifying the type of leak, maintenance teams can tailor their repair strategies to address the specific issue, thus minimizing downtime and resource usage. Different types of leaks pose varying levels of risk to the pipe system's integrity; for instance, a circumferential break can lead to catastrophic failure and significant fluid loss. The apparatus and method disclosed herein allow for more accurate risk assessments and prioritization of repairs, ensuring critical issues are addressed promptly. After switching between the sensing modes and / or varying the characteristics of the emitted light pulses, the processing unit may be configured to process the backscattering signal detected by light detector module at a second point in time. The second point in time may be a point in time after the sensing modes have been switched and / or after the characteristics of the emitted light pulses have been varied. By processing the backscattering signal detected by light detector module at a second point in time, the processing unit may be able to classify the anomalous event or the inconclusive event as the particular event.

[0024] This subsequent analysis is crucial because the initial detection might have produced ambiguous or inconclusive data, necessitating further investigation. By altering the sensing modes and pulse characteristics, the system gathers additional, more refined data that provides a clearer picture of the event. At this second point in time, the enhanced backscattering signal, now enriched with complementary information, enables the processing unit to more accurately classify the anomalous or inconclusive event.

[0025] Upon not being able to classify the anomalous event or the inconclusive event at the second point in time, the processing unit may be configured to continue processing the backscattering signal until a third point in time. The third point in time may be after the second point in time. Additionally, the processing unit may be configured to average the backscattering signal over a period of time between the second point in time and the third point in time to output an averaged backscattering signal. Subsequently, the processing unit may be configured to use the averaged backscattering signal to classify the anomalous event or the inconclusive event as the particular event.

[0026] Averaging the backscattering signal over a period of time between the second and third points in time offers several advantages in accurately classifying an anomalous or inconclusive event. When initial and subsequent analyses fail to conclusively identify the event, continuous processing and averaging the signal helps to mitigate the effects of transient noise and random fluctuations. This approach enhances the signal-to-noise ratio, making persistent and consistent features of the signal more prominent while reducing the impact of sporadic or erratic disturbances. By smoothing out short-term variances, the averaged backscattering signal provides a clearer and more stable representation of the underlying event. This improved clarity allows the processing unit to detect subtle patterns and characteristics that might not be evident in a single snapshot or short-term analysis.

[0027] The processing unit may be configured to generate an alarm in response to classifying the anomalous event or the inconclusive event as a leak in the pipe, the pipe bursting, a structural deformation of the pipe or a blockage in the pipe.

[0028] In some embodiments, the alarm may comprise data indicative of the position of the leak in the pipe. In some embodiments, the alarm may additionally include data indicative of the size or type of leak or of the event. In some embodiments, the alarm signal may be generated in response to identification of an acoustic candidate using a single detection mode. In further embodiments, the alarm signal may only be generated in response to identification of a leak candidate using multiple different sensing modes. In some embodiments, the processing unit may be configured to compare the position of the respective leak candidates detected using different sensing modes to ensure they are in the same position or within a preset displacement form each other before the alarm signal can be generated.

[0029] Advantageously, by promptly detecting and alerting operators to critical events, the system enables rapid response measures to be implemented, minimizing potential damage and disruption to operations. For instance, in the case of a bust pipe, immediate notification allows for the swift isolation of affected pipe sections from the network and the initiation of emergency repair protocols, preventing extensive flooding or loss of fluid from the pipe network.

[0030] The multiple different sensing modes may comprise a distributed acoustic sensing (DAS) mode. The multiple different sensing modes may comprise a distributed strain sensing (DSS) mode. The multiple different sensing modes may comprise a distributed temperature sensing (DTS) mode.

[0031] DAS mode may comprise a sensing technique utilized within the monitoring apparatus to detect acoustic signals and vibrations along the length of the sensing fibre. In the DAS mode, Rayleigh scattering within the sensing fibre may be leveraged for sensing purposes.

[0032] The light detector module may be configured to detect the backscattering of the light pulses using the DAS mode. The light detector module may be configured to output a DAS backscattering signal in response to detecting the backscattering of the light pulses using the DAS mode. In some embodiments, the DAS backscattering signal may be labelled as a Rayleigh detection signal. The processing unit may be configured to process the DAS backscattering signal to determine an acoustic vibration experienced by particular points on the sensing fibre, and hence particular locations along the pipe. The processing unit may be configured to compare the determined acoustic vibration against expected acoustic profiles of particular events to identify the particular event.

[0033] More specifically, Rayleigh detection signal may be processed to determine vibration amplitudes and frequencies experienced by particular scattering points on the sensing fibre and hence particular locations along the pipe. In some embodiments, the received Rayleigh detection signal may be filtered. The filtering may be in respect of time of receipt (and hence location along the sensing fibre) or in respect of vibration frequency, vibration amplitude, or a combination thereof. In particular, determined vibration frequency and amplitude may be compared against expected signatures of particular events. This can enable detection or exclusion of particular sources of vibration. In one example, this could involve matching Rayleigh detection signals to preset orifice noise profiles, so as to identify the particular event. In other examples, this could involve excluding processing of signals with a profile characteristic of external noise sources to the pipe. This might thereby exclude traffic on roads overlying or close to the pipe.

[0034] DSS mode may comprise a sensing technique employed within the monitoring apparatus to detect static and quasi-static strain variations along the sensing fibre (and thus the fluid pipe). DSS may utilise Brillouin scattering for sensing purposes. In DSS, light pulses injected into the sensing fibre may undergo Brillouin scattering, and the resulting frequency shift may be measured to determine strain variations along the fibre. By analysing these shifts in the backscattered light, DSS can accurately detect and localize strain-induced events, such as structural deformations, pressure changes, or mechanical stresses, along the entire length of the fibre (and thus the associated length of the pipe comprising the fibre).

[0035] The light detector module may be configured to detect the backscattering of the light pulses using the DSS mode. The light detector module may be configured to output a DSS backscattering signal in response to detecting the backscattering of the light pulses using the DSS mode. The processing unit may be configured to process the DSS backscattering signal to determine strain induced variations experienced by particular points on the sensing fibre, and hence particular locations along the pipe. The processing unit may be configured to compare the determined strain induced variations against expected strain profiles of particular events to identify the particular event.

[0036] DTS mode may comprise an integral sensing technique utilized by the monitoring apparatus to measure temperature variations along the sensing fibre. In DTS, Brillouin scattering and / or Raman scattering may be used for sensing purposes. When light pulses are injected into the fibre, these pulses may interact with the fibre’s molecular structure, undergoing Brillouin and Raman scattering. The resulting frequency shifts and intensity changes may be measured to determine temperature variations along the fibre. By analysing these changes in the backscattered light, temperature fluctuations, such as hotspots, thermal gradients, or ambient temperature changes, may be detected and localized along the entire length of the fibre (and thus along the entire length of the fluid pipe).

[0037] The light detector module may be configured to detect the backscattering of the light pulses using the DTS mode. The light detector module may be configured to output a DTS backscattering signal in response to detecting the backscattering of the light pulses using the DTS mode. The DTS backscattering signal may be labelled as a Brillouin detection signal and / or a Raman detection signal. The processing unit may be configured to process the DTS backscattering signal to determine temperature variations experienced by particular points on the sensing fibre, and hence particular locations along the pipe. The processing unit may be configured to compare the determined temperature variations against expected temperature profiles of particular events to identify the particular event. Brillouin detection signals and / or Raman detection signals may be processed to determine temperature variations at particular locations along the pipe. The determined temperature variations may be compared against expected temperature profiles of particular events. This can enable detection or exclusion of particular sources of temperature change. In one example, this could involve matching a temperature change profile to an adiabatic cooling signature, so as to identify a leak. In other examples, this could involve excluding processing of signals with a profile characteristic of a known external heat source / sink such as an industrial site overlying or close to a buried pipe.

[0038] In some embodiments, the light detector module may be configured to initially detect backscattering of the light pulses from the sensing fibre using the DAS mode. Additionally, the processing unit may be configured to control the light detector module to switch to the DSS mode or the DTS mode upon detecting the anomalous event or the inconclusive event.

[0039] This monitoring approach is especially advantageous in scenarios involving leak detection. For instance, if DAS mode initially detects an acoustic profile suggestive of a potential small leak, but the detected profile remains inconclusive, the system can seamlessly transition to DTS mode to further investigate. Temperature modification in liquid pipes may result from interaction with surrounding ground conditions. By monitoring for adiabatic temperature changes in gaseous fluid, such as adiabatic cooling, along the length of the pipe, DTS provides additional data points to corroborate the presence of a leak. The detection of a temperature change in conjunction with the acoustic profile identified by DAS enhances confidence that a leak is indeed present. This approach may be able to successfully identify even small leaks in the pipe.

[0040] Upon detecting the anomalous event or the inconclusive event, the processing unit may be configured to vary the characteristics of the emitted light pulses. Subsequently, the processing unit may be configured to determine whether varying the characteristics of the emitted light pulses leads to a successful classification of the anomalous event as the particular event. Additionally, or alternatively, the processing unit may be configured to determine whether varying the characteristics of the emitted light pulses leads to a successful classification of the inconclusive event as the particular event. Upon determining that varying the characteristics of the emitted light pulses does not lead to a successful classification of the anomalous event or the inconclusive event, the processing unit may be configured to cause the light detector module to switch between the sensing modes.

[0041] Varying the characteristics of the emitted light pulses may comprise reducing or increasing a pulse width. Varying the characteristics of the emitted light pulses may comprise reducing or increasing a pulse repetition frequency. Varying the characteristics of the emitted light pulses may comprise reducing or increasing a gauge length.

[0042] Increasing the pulse width may enhance sensitivity by allowing for more photons to interact with the fibre, thereby increasing the signal-to-noise ratio and improving the detection of subtle changes in the backscattered light. However, this increase in sensitivity may come at the expense of spatial resolution, as longer pulse widths may result in reduced ability to pinpoint the location of events along the fibre. Conversely, decreasing the pulse width may enhance spatial resolution enabling finer localization of events, albeit at the cost of sensitivity due to fewer photons being available for interaction with the fibre. By dynamically adjusting the pulse width based on the specific monitoring requirements and environmental conditions, the apparatus may be able to adapt to optimize either sensitivity or spatial resolution as needed.

[0043] Increasing the pulse repetition frequency may improve the signal-to-noise ratio of the backscattering signal. Increasing the pulse repetition frequency may increase the rate at which light pulses are emitted and detected, resulting in a greater number of measurements over a given time interval and thus a better signal-to-noise ratio. The DAS system may respond to any acoustic frequency on the fibre, but due to Nyquist sampling criteria, only acoustic frequencies up to half the pulse repetition frequency may be classified correctly by the apparatus. As such, the pulse repetition frequency may be selected based on: the length of the sensing fibre in the system, the typical audio bandwidth associated with a particular event (e.g., a leak) and the required signal-to- noise ratio.

[0044] The gauge length may comprise a length of the sensing fibre over which the measurements are averaged. The measurement for each gauge length may comprise an average of the backscattering signal from multiple light pulses. This averaging of the backscattering signal from multiple light pulses may improves the signal-to-noise ratio of the backscattering signal. As such, increasing the gauge length may improve the signal-to-noise ratio of the backscattering signal. However, increasing the gauge length may make it more difficult to determine the exact location of the particular event along the length of the pipe. More specifically, the broader averaging window provided by a larger gauge length may results in a loss of spatial resolution, making it more difficult to pinpoint the precise location of events such as leaks, structural deformations, or blockages.

[0045] In some embodiments, varying the characteristics of the emitted light pulses may comprise starting with a light pulse having a broad gauge length to identify the particular event and subsequently narrowing the gauge length to identify the precise location of the particular event.

[0046] Detecting the inconclusive event may comprise assigning a confidence level to each of the identified backscattering signal features within the backscattering signal. Detecting the inconclusive event may comprise determining that the confidence level of a first backscattering signal feature or a first group of backscattering signal features lies in an intermediate confidence range. Detecting the inconclusive event may comprise determining that the first backscattering signal feature or the first group of backscattering signal features are indicative of an inconclusive event.

[0047] The inconclusive confidence range may serve as a crucial threshold within the monitoring apparatus, delineating levels of certainty regarding the presence or absence of particular events. Positioned between a low and a high confidence level, this range may represent a zone of ambiguity where the interpretation of backscattering signal features is neither definitively affirmative nor outright negative.

[0048] More specifically, the inconclusive confidence range may lie between a low confidence level and a high confidence level. Below the low confidence level, the processing unit may be configured to recognise that no particular event has taken place. For example, below the low confidence level, the processing unit may be configured to recognise that there is no leak within the pipe. Above the high confidence level, the processing unit may be configured to immediately recognise that the particular event has taken place. For example, above the high confidence level, the processing unit may be configured to recognise that there is indeed a leak within the pipe (without having to switch the sensing modes or vary the characteristics of the light pulses).

[0049] Detecting the anomalous event may comprise monitoring the backscattering signal for a first period of time to detect typical backscattering signal features associated with the pipe or a portion of the pipe. Detecting the anomalous event may comprise processing the backscattering signal at a point in time after the first period of time to identify a second group of backscattering signal features. Detecting the anomalous event may comprise comparing the second group of backscattering signal features to the typical backscattering signal features. Detecting the anomalous event may comprise determining that that the second group of backscattering signal features differs from the typical backscattering signal features. Lastly, detecting the anomalous event may additionally comprise determining that the second group of backscattering signal features are indicative of the anomalous event.

[0050] Advantageously, the process of detecting anomalous events by the processing unit involves a comprehensive approach aimed at distinguishing abnormal signal features from typical ones associated with the pipe or its operation. By initially monitoring the backscattering signal over a predefined period, the apparatus may establish a baseline of typical signal characteristics. Subsequently, processing the signal at a later time, may allow for the identification of deviations or anomalies, which may indicate potential issues or events requiring further investigation. Even if the detected signal features surpass the high confidence level, further scrutiny may be warranted to accurately identify the specific event.

[0051] In one example, backscattering signal features indicative of a large leak may be detected using the DAS mode, but no preceding buildup or warning signs may be observed by the apparatus. The large leak (in view of the lack of prior warning signs) may comprise an anomalous event. The apparatus may switch to alternative sensing modes, such as DTS or DSS. In this embodiment, after switching the sensing modes, the apparatus may conclude that the detected backscattering signal features are not indicative of a leak but rather an opening of a valve. As such, the apparatus and method disclosed herein may accurately characterize or classify anomalous events, thereby minimizing false alarms. Determining that that the second group of backscattering signal features differs from the typical backscattering signal features may comprise determining that the difference between characteristics of the second group of backscattering signal features and the typical backscattering signal features is above a pre-defined threshold.

[0052] Characteristics of the backscattering signal feature may comprise one or more of: signal intensity of the backscattering signal feature or features, signal frequency of the backscattering signal feature or features, signal amplitude of the backscattering signal feature or features etc. The skilled person would easily recognise other characteristics of the backscattering signal features which may be compared when comparting the second group of backscattering signal features and the typical backscattering signal features.

[0053] If, during monitoring, the second group of signal features displays a sudden spike in intensity or a pronounced shift in frequency beyond the predefined threshold, this may indicate the presence of an anomaly, such as a sudden pressure surge or mechanical disturbance within the pipe. Similarly, deviations exceeding the threshold could manifest as irregularities in the temporal or spatial distribution of backscattered light, suggesting potential structural deformations or blockages. By establishing clear criteria for distinguishing between typical and anomalous signal characteristics, the monitoring apparatus can effectively flag events requiring attention, enabling timely intervention and maintenance to ensure the integrity and functionality of the fluid pipe system.

[0054] In some embodiments, upon detecting an anomalous event or an inconclusive event the processing unit may be configured to process the backscattering signal using a processing algorithm. The processing algorithm may be configured to extend the dynamic range of the backscattering signal. The processing algorithm may improve the confidence level of each of the identified backscattering signal features within the backscattering signal. By improving the confidence level of each of the identified backscattering signal features within the backscattering signal, the processing algorithm may be configured to classify the anomalous event or the inconclusive event as a particular event. In some embodiments, the processing algorithm may only be employed after switch between the sensing modes and / or varying the characteristics of the emitted light pulses has failed to classify the anomalous event or the inconclusive event as a particular event.

[0055] This selective use of the processing algorithm offers several advantages. Primarily, it conserves processing resources by avoiding the immediate deployment of a highly processor- intensive algorithm, thereby optimizing the apparatus’s overall efficiency. By first attempting to classify the event through simpler, less resourcedemanding methods — such as switching sensing modes or adjusting light pulse characteristics — the apparatus prioritizes more efficient pathways to event classification.

[0056] The sensing fibre may be an optical fibre. The optical fibre may be adapted to carry or preferentially carry any suitable wavelength of light. Suitable wavelengths may be wavelengths emitted by the light emitter.

[0057] The sensing fibre may be a single fibre. The sensing fibre may be a dedicated fibre within a bundle of fibres. The bundle of fibres may form a multicore fibre cable. The DAS monitoring preferably operates from one end only. As such the sensing fibre may be a single ended fibre.

[0058] Preferably a barrier is provided between the sensing fibre and the fluid. This can protect the sensing fibre from damage from the fluid or debris within the fluid. The barrier may comprise a coating or cover provided over the sensing fibre. In the event that the sensing fibre is provided within a multicore fibre, the barrier may comprise a coating or cover provided over the multicore fibre. Additionally or alternatively, the barrier may comprise a microduct within which the fibre is provided. In such embodiments, the method may include the step of introducing the microduct to the pipe and subsequently blowing the sensing fibre along the microduct.

[0059] In the event that there is a gap between the sensing fibre and the barrier, the gap may be filled with gel. This can improve acoustic coupling between the sensing fibre and the pipe or fluid within the pipe. The method may include the step of introducing a gel between the sensing fibre and the barrier. The sensing fibre may be fixedly connected or mounted to an inside surface of the pipe. By mounting the sensing fibre to the inside surface of the pipe, DSS monitoring may be successfully used measure strain experienced by the fluid pipe.

[0060] According to a second aspect of the invention, there is provided a method for monitoring a fluid pipe. The method may comprise providing at least one sensing fibre within the pipe. The method may comprise introducing light pulses with particular characteristics into the fibre using a light emitter. The method may comprise detecting backscattering of the light pulses from the sensing fibre using a light detector module operating in a first sensing mode. The method may comprise outputting, by the light detector module, a backscattering signal in response to the detected backscattering of the light pulses. The method may comprise processing, by a processing unit, the backscattering signal to identify backscattering signal features characteristic of a particular event. The method may comprise detecting, by the processing unit, an anomalous event or an inconclusive event. The method may comprise in repose to detecting the anomalous event or the inconclusive event causing the light detector module to switch to a different sensing mode. The method may comprise in repose to detecting the anomalous event or the inconclusive event causing the light emitter to vary the characteristics of the emitted light pulses.

[0061] In some embodiments, the first sensing mode may comprise a distributed acoustic sensing (DAS) mode and the different sensing mode may comprise either a distributed strain sensing (DSS) mode or a distributed temperature sensing (DTS) mode. The skilled person will recognise that in some embodiments, the first sensing mode may comprise any one of DAS, DTS or DSS. The skilled person will recognise that in some embodiments, the different sensing mode may comprise any one of DAS, DTS or DSS.

[0062] Varying the characteristics of the emitted light pulses may comprise reducing or increasing one or more of: a pulse width, a pulse repetition frequency; and / or a gauge length.

[0063] After switching to the different sensing mode and / or varying the characteristics of the emitted light pulses, the method may comprise processing, by the processing unit, the backscattering signal detected by light detector module at a second point in time. Subsequently, the method may further comprise classifying the anomalous event or the inconclusive event as a particular event.

[0064] The method may further comprise generating, by the processing unit, an alarm in response to classifying the anomalous event or the inconclusive event as a leak in the pipe, the pipe bursting, a structural deformation of the pipe or a blockage in the pipe.

[0065] According to a third aspect of the present invention, there is provided a pipe network comprising one or more pipes monitored using the apparatus of the first aspect of the present invention and / or the method of the second aspect of the present invention.

[0066] The skilled person will appreciate that except where mutually exclusive, a feature described in relation to any one of the aspects, methods, examples or embodiments described herein may be applied to any other method, aspect, example, embodiment or feature. Further, the description of any aspect, method, example or feature may form part of or the entirety of an embodiment of the invention as defined by the claims. Any of the examples described herein may be an example which embodies the invention defined by the claims and thus an embodiment of the invention.

[0067] Detailed Description of the Invention

[0068] In order that the invention may be more clearly understood one or more embodiments thereof will now be described, by way of example only, with reference to the accompanying drawings, of which:

[0069] Figure la is a schematic block diagram of a sensing apparatus operable to carry out distributed acoustic sensing (DAS), distributed strain sensing (DSS) and distributed temperature sensing (DTS);

[0070] Figures lb- Id show alternative embodiments of a light detector module for use in the sensing apparatus of figure la;

[0071] Figure 2 is (a) a schematic illustration of pipe monitoring using a sensing fibre according to the present invention and (b) a schematic illustration of temperature variation along the pipe of figure 2a;

[0072] Figure 3a is a schematic illustration of a leak in a pipe. Figure 3b is a schematic illustration of acoustic variation along the pipe of figure 3a.

[0073] Figure 3c is a schematic illustration of a gauge length of an emitted light pulse.

[0074] Figure 3d is a schematic illustration of an emitted light pulse having a certain pulse width and pulse repetition frequency.

[0075] Figure 4 is a schematic illustration of a backscattering signal detected at a light detector module.

[0076] Figure 5a-b show a flow diagram of a method for monitoring a fluid pipe.

[0077] Figure 6 is (a) a flow diagram of a method for detecting an anomalous event and

[0078] (b) a flow diagram of a method for detecting an inconclusive event.

[0079] Figure la shows a schematic block diagram of a sensing apparatus 100 operable to carry out distributed acoustic sensing (DAS), distributed strain sensing (DSS) and distributed temperature sensing (DTS) in fluid pipe 1 (best seen in Figure 2a). The fluid pipe 1 is monitored using a sensing fibre 10. In the description below, the invention is primarily described in terms of monitoring of a pipe carrying water. Nevertheless, the skilled person will appreciate that the invention may be applied to pipes carrying other fluids including, but not limited to, waste water, sewage or fuels such as oil, gas distillates or the like and chemical or mining products.

[0080] The sensing fibre is an optical fibre coupled at one end to a light emitter 101, typically a laser and also to a light detector module 110.

[0081] The light emitter 101 emits light pulses into the sensing fibre. The light pulses have particular characteristics as will be described in more detail with reference to Figures 3c and 3d.

[0082] The light detector module 110 detects backscattered light from the sensing fibre 10 and the time of arrival of the backscattered light following the emission of a pulse relates to the location of the backscattering site along the sensing fibre 10. The light detector module 110 is configured to detect backscattering of the light pulses from the sensing fibre 10 in multiple different sensing modes and output a backscattering signal in response thereto. More specifically, the light detector module 110 is configured to detect backscattering of the light pulses from the sensing fibre in the distributed acoustic sensing (DAS) mode, the distributed strain sensing (DSS) mode and the distributed temperature sensing (DTS) mode. The light detector module 110 is configured to output a DAS, DTS or DSS backscattering signal in response to detecting the backscattering of the light pulses from the sensing fibre in a DAS mode, a DTS mode or the DSS mode, respectively.

[0083] DAS mode comprises a sensing technique utilized within the monitoring apparatus 100 to detect acoustic signals and vibrations along the length of the sensing fibre 10. In the DAS mode, Rayleigh scattering within the sensing fibre 10 is leveraged for sensing purposes. As such, the DAS mode may be labelled as a Rayleigh scattering mode and the DAS backscattering signal may be labelled as a Rayleigh detection signal.

[0084] DSS mode comprises a sensing technique employed within the monitoring apparatus 100 to detect static and quasi-static strain variations along the sensing fibre 10. DSS utilises Brillouin scattering for sensing purposes. In DSS, light pulses injected into the sensing fibre 10 by the light emitter 101 undergo Brillouin scattering, and the resulting frequency shift is measured to determine strain variations along the fibre. As such, the DSS mode may be labelled as a Brillouin scattering mode and the DSS backscattering signal may be labelled as a Brillouin detection signal.

[0085] DTS mode comprises an integral sensing technique utilized by the monitoring apparatus 100 to measure temperature variations along the sensing fibre 10. In DTS, Brillouin scattering and / or Raman scattering are used for sensing purposes. As such, the DTS mode may be labelled as Brillouin scattering and / or Raman scattering mode and the DTS backscattering signal may be labelled as a Brillouin scattering and / or Raman scattering detection signal.

[0086] In some embodiments, a single backscattering signal captured by the light detector module 110 comprises a signal feature indicative of Rayleigh scattering, a signal feature indicative of Brillouin scattering and a signal feature indicative of Raman scattering. The signal feature indicative of Rayleigh scattering may be analysed in the DAS mode, the signal feature indicative of Brillouin scattering may be analysed during the DSS or the DTS mode and a signal feature indicative of Raman scattering may be analysed during DTS. This example will be described in more detail with reference to Figure 4.

[0087] The detected DAS backscattering signal (i.e., Rayleigh detection signal indicative of Rayleigh scattering within the sensing fibre 10), the detected DSS backscattering signal (i.e., Brillouin detection signal indicative of Brillouin scattering within the sensing fibre 10) and / or the detected DTS backscattering signal (i.e., Brillouin scattering and / or Raman scattering detection signals indicative of Brillouin scattering and / or Raman scattering within the sensing fibre 10) are passed to a processing unit 103 for processing.

[0088] The processing unit 103 will typically be local to the light emitter 101 and light detector 110 but may be alternatively provided at a remote location. In the latter case, a communication unit (not shown) would be operable to communicate remotely with the processing unit 103.

[0089] The light detector module 110 may be a combined Rayleigh spectrometer and Brillouin spectrometer and / or Raman spectrometer as illustrated in Figure la. In this embodiment, the light detector module 110 may comprise a multimode spectrometer.

[0090] The Rayleigh spectrometer may be configured to detect the backscattering of the light pulses from the sensing fibre 10 in the DAS mode and output a DAS backscattering signal in response thereto. The skilled person would understand that Rayleigh spectrometer may be called a DAS spectrometer. The Brillouin spectrometer may be configured to detect the backscattering of the light pulses from the sensing fibre 10 in the DSS mode and output a DSS backscattering signal in response thereto. The skilled person would understand that the Brillouin spectrometer may be called a DSS spectrometer. The Brillouin spectrometer and / or a Raman spectrometer may be configured to detect the backscattering of the light pulses from the sensing fibre 10 in the DTS mode and output a DTS backscattering signal in response thereto. The skilled person would understand that the Brillouin spectrometer and / or the Raman spectrometer may be called a DTS spectrometer.

[0091] Alternatively, the light detector module 110 may comprise a Rayleigh spectrometer 112 and a separate Brillouin spectrometer 113 only, as illustrated in Figure lb. In another alterative, the light detector module 110 may comprise a Rayleigh spectrometer 112 and a separate Raman spectrometer 114 only, as illustrated in Figure lc. In a further alternative, the light detector module 110 may comprise a Rayleigh spectrometer 112 and a separate Brillouin spectrometer 113 as well as a separate Raman spectrometer 114, as illustrated in Figure Id.

[0092] In embodiments having separate spectrometers, the light detector module 110 can also comprise a light guiding device 111 configured to direct backscattered light to Rayleigh spectrometer 112 and either spectrometer or both spectrometers 113, 114 as appropriate. The light guiding device may be a beam splitter. Otherwise, the light guiding device may be a switchable light guiding device configured to direct light selectively to the Rayleigh spectrometer 112 or the Brillouin spectrometer 113 and / or the Raman spectrometer 114.

[0093] The light detector module 110 may be configured to switch between the DAS, DTS and DSS modes. In one example, the light detector module 110 is configured to operate primarily in the DAS mode and the processing unit 103 may switch to DTS or DSS mode in response to detecting an anomalous event or an inconclusive event as will be described with reference to Figures 5b, 6a and 6b. In another example, the processing unit 103 is configured to cause the light emitter 101 to vary the characteristics of the emitted light pulses in response to detecting an anomalous event or an inconclusive event as will be described with reference to Figures 5b, 6a and 6b.

[0094] In more detail, the light detector module 110 is configured to detect the backscattering of the light pulses using the DAS mode (e.g., by detecting one or more signal feature indicative of Rayleigh scattering). The light detector module 110 is then configured to output a DAS backscattering signal (i.e., a Rayleigh detection signal) in response to detecting the backscattering of the light pulses using the DAS mode. The processing unit 103 is subsequently configured to process the DAS backscattering signal to determine an acoustic vibration experienced by particular points on the sensing fibre 10, and hence particular locations along the pipe 1. The processing unit 103 is configured to compare the determined acoustic vibration against expected acoustic profiles of particular events to identify the particular event. The particular event could be a leak in the pipe 1 , a change in external conditions proximal to the pipe 1, the pipe 1 bursting, a structural deformation of the pipe 1, a change in fluid flow within the pipe 1 , a change in fluid pressure within the pipe 1 , a blockage in the pipe 1, and / or a valve connected to the pipe 1 being opened or closed.

[0095] The processing unit 103 can be configured to compare the determined vibration frequency and amplitude from the DAS backscattering signal against expected vibration profiles of particular events. This can enable detection or exclusion of particular sources of vibration. In one example, this could involve matching the DAS backscattering signals to preset orifice noise profiles, flow noise profiles and / or negative pressure waves indicative of leaks. In this way, the processing unit 103 can identify potential leak candidates. Conversely, the processing unit 103 can excluding processing of signals with a profile characteristic of external noise sources to the pipe. This might thereby exclude traffic noise from the locality of a buried pipe.

[0096] In some embodiments, the light detector module 110 may be configured to detect the backscattering of the light pulses using the DSS mode (e.g., by detecting one or more signal feature indicative of Brillouin scattering). In this embodiment, the light detector module 110 is configured to output a DSS backscattering signal in response to detecting the backscattering of the light pulses using the DSS mode. The processing unit 103 is subsequently configured to process the DSS backscattering signal to determine strain induced variations experienced by particular points on the sensing fibre 10, and hence particular locations along the pipe 1. The processing unit 103 is configured to compare the determined strain induced variations against expected strain profiles of particular events (e.g., leaks) to identify the particular event.

[0097] In particular, as the sensing fibre 10 may be mounted to the pipe 1 (as shown in Figure 2a) and DSS can be carried out to assess the overall strain on pipe 1 and / or joints between pipe (not shown). The pipe 1 may experience such strain in response to variations in internal or external loading. The internal loading can result from variations in flow volume, flow rate or the like. The external loading may result from shifting of earth surrounding a buried pipe 1 , shifting of the supports for an above ground pipe or the like. In some embodiments, the light detector module 110 may be configured to detect the backscattering of the light pulses using the DTS mode (e.g., by detecting one or more signal feature indicative of Brillouin and / or Raman scattering). The light detector module 110 is subsequently configured to output a DTS backscattering signal in response to detecting the backscattering of the light pulses using the DTS mode. The processing unit 103 is subsequently configured to process the DTS backscattering signal to determine temperature variations experienced by particular points on the sensing fibre 10, and hence particular locations along the pipe 1. The processing unit 103 is subsequently configured to compare the determined temperature variations against expected temperature profiles of particular events (e.g., leaks) to identify the particular event.

[0098] In one example, the processing unit 103 may match the DTS backscattering signal to a temperature change profile indicative of adiabatic cooling. Such cooling may occur due to pressure loss at a leak site. Additionally or alternatively, such cooling may occur due to fluid egress into a surrounding substrate. As such, identifying such a temperature reduction may enable identification of a potential leak. In other examples, this could involve excluding processing of signals with a profile characteristic of a known external heat source / sink such as an industrial site in the locality of a buried pipe 1.

[0099] The processing unit 103 is configured to generate an alarm in response to identifying a particular event as a leak in the pipe 1, the pipe 1 bursting, a structural deformation of the pipe 1 or a blockage in the pipe 1.

[0100] In some embodiments, the alarm may comprise data indicative of the position of the leak in the pipe 1. In some embodiments, the alarm may additionally include data indicative of the size or type of leak. In some embodiments, the alarm signal may be generated in response to identification of a leak candidate using the DAS detection mode. To reduce false alarms and increase certainty, the alarm signal may only be generated in response to identification of a leak candidate using multiple different sensing modes (e.g., DAS, DTS and / or DSS sensing modes). In some embodiments, the processing unit 103 is configured to compare the position of the respective leak candidates detected using different sensing modes (DAS, DTS and DSS) to ensure they are in the same position or within a preset displacement form each other before the alarm signal is generated.

[0101] Turning now to figure 2a, the sensing fibre 10 is provided within pipe 1. In the event that the pipe 1 has a leak 2, vibrations 3 characteristic of orifice noise will travel through the fluid until they impinge on the sensing fibre 10. Such vibration cause Rayleigh scattering and thus provide the DAS backscattering signal which is detected by the light detection module 110. Subsequent operation of the processing unit 103 will determine the occurrence of vibrations of the sensing fibre 10 and the position along the sensing fibre 10 at which these vibrations occur. Accordingly, the processing unit 103 can identify a leak 2 and the position of the leak 2 along the length of pipe 1.

[0102] As shown in figure 2a, the pipe 1 runs under a road 20. Vehicles 21 travelling along the road 20 generate vibrations 22 which can travel though the ground to the pipe 1. The vibrations 22 generate additional DAS backscattering signals that are also detected on sensing fibre 10. If the leak 2 is small (for example, if the leak 2 comprises a relatively small pinhole leak), the DAS backscattering signal due to the leak 2 may be difficult to detect and classify correctly. Even if the leak 2 is larger, the DAS backscattering signal due to the leak 2 may be partially masked by the DAS backscattering signals generated by the vibrations 22. In these embodiments, the processing unit 103 may determine that a DAS backscattering signal feature or a group of DAS backscattering signal features within the DAS backscattering signal detected by the light detector module 110 are indicative of an inclusive event. Upon identifying an inconclusive event, the processing unit 103 employs various techniques to try and classify the inconclusive event as the particular event (e.g., a leak 2 in the pipe 1). These techniques will be described in more detail with reference to Figures 5 a, 5b and 6b.

[0103] Turning now to figure 2b, the temperature profile of the pipe fluid generated from analysis of the DTS detection signal is also illustrated. As illustrated, the temperature profile has a dip 15, which is characteristic of adiabatic cooling associated with a leak from a pressurised fluid pipe. Processing of the DTS signal by the processing unit 103 can therefore confirm that an inconclusive event detected using the DAS mode indeed comprises a leak 2. The DTS detection signal can also help the processing unit 103 to more accurately identify a position of the leak 2 along the length of pipe 1. Although not shown in the Figures, there may exist a strain profile of the pipe 1 generated from analysis of the DSS detection signal. The strain profile of the pipe 1 may also indicate a presence of the leak 1. Processing of the DSS signal by the processing unit 103 can therefore confirm that an inconclusive event detected using the DAS mode indeed comprises a leak 2. The DSS detection signal can also help the processing unit 103 to identify a position of the leak 2 along the length of pipe 1.

[0104] In the example shown, the sensing fibre 10 is a single dedicated fibre in a multicore cable 11 formed from a plurality of fibres. The sensing fibre 10 is mounted to an inside surface of the pipe 2. The multicore cable 11 may be provided with a protective exterior coating (not shown). Furthermore, in this example, the multicore cable 11 is provided within a microduct 12. The microduct 12 forms a barrier between the cable 11 and the fluid within pipe 1. In the description below, the invention will primarily be described in terms of dedicated sensing fibres 10 of multicore cables 11 provided within microducts 12. Nevertheless, the skilled person will appreciate that the invention may be applied to multicore cables 11 or single sensing fibres 10 (both either coated or uncoated) provided directly within pipes rather than within a microduct 12. The microduct 12 may be filled with acoustic gel or fluid (not shown). This can improve acoustic coupling between the microduct 12 and the sensing fibre 10.

[0105] Turning now to Figure 3 a there is shown a schematic illustration of the pipe 1 comprising the leak 2. Figure 3b shows a gaussian power distribution profile 306 of the vibrations 3 generated by the leak 2. In this figure, the x-axis 304 shows distance along the length of the pipe 1 and the y-axis 302 shows amplitude of the acoustic backscattering signal detected by the light detector module 110 operating in the DAS mode.

[0106] The acoustic profile 306 comprises a substantially gaussian profile, wherein a maximum 308 amplitude of the acoustic signal is indicative of a position of the leak 2 along the pipe 1. By identifying an x-axis position of the maxima 308 of the acoustic signal, the processing unit 103 is able to determine the position of the leak 2 along the pipe 1. By analysing the acoustic profile 306, the processing unit 103 is able to distinguish between different types of leaks in the pipe, including pinhole leaks, longitudinal cracks, and circumferential breaks.

[0107] The processing unit 103 is configured to control the light emitter 101 to vary the characteristics of the emitted light pulses. Characteristics of the light pulses are schematically represented in Figures 3c and 3d.

[0108] Figure 3c shows a schematic illustration of an emitted light pulse 330 having a gauge length 332. Figure 3d shows a schematic illustration of the emitted light pulse 330 which has a pulse repetition frequency 342 and a pulse width 344.

[0109] Varying the characteristics of the emitted light pulse 330 may comprise reducing or increasing: the pulse width 344, the pulse repetition frequency 342 and / or the gauge length 332.

[0110] Increasing the pulse width 344 enhances sensitivity by allowing for more photons to interact with the fibre 10, thereby increasing the signal-to-noise ratio and improving the detection of subtle changes in the backscattered light. However, this increase in sensitivity comes at the expense of spatial resolution, as longer pulse widths 344 result in reduced ability to pinpoint the location of the leak 2 (or other events) along the fibre 10. Conversely, decreasing the pulse width 344 enhances spatial resolution enabling finer localization of the leak 10 (or other events), albeit at the cost of sensitivity due to fewer photons being available for interaction with the fibre 10. By dynamically adjusting the pulse width 344, the processing unit 103 may be able to adapt to optimize either sensitivity or spatial resolution as needed, thus allowing the processing unit 103 to classify any anomalous events or the inconclusive events as a particular event (e.g., a leak 2), as will be described with reference to Figures 5b, 6a and 6b.

[0111] Increasing the pulse repetition frequency 342 improves the signal-to-noise ratio of the backscattering signal 330. Increasing the pulse repetition frequency 342 increases the rate at which light pulses are emitted and detected, resulting in a greater number of measurements over a given time interval and thus a better signal-to-noise ratio. The Rayleigh spectrometer 112 may respond to any acoustic frequency on the fibre 10, but due to Nyquist sampling criteria, only acoustic frequencies up to half the pulse repetition frequency may be classified correctly by the apparatus. As such, the pulse repetition frequency may be selected based on: the length of the sensing fibre 10, the typical audio bandwidth associated with a particular event (e.g., a leak) and the required signal-to-noise ratio. By dynamically adjusting the pulse frequency, the processing unit 103 may be able to improve the signal-to-noise ratio in the detected backscattering signal, thus allowing the processing unit 103 to classify any anomalous events or the inconclusive events as a particular event (e.g., a leak 2), as will be described with reference to Figures 5b, 6a and 6b.

[0112] The gauge length 332 comprises a length of the sensing fibre 10 over which the measurements are averaged. In this schematic example, the gauge length 332 comprises an average of the backscattering signal from five light pulses. This averaging of the backscattering signal from multiple light pulses improves the signal-to-noise ratio of the backscattering signal. As such, increasing the gauge length 332 improves the signal- to-noise ratio of the backscattering signal. However, increasing the gauge length 332 makes it more difficult for the processing unit 103 to determine the exact location of the maxima 308 of the acoustic signal 306 (and thus the location of the leak 2) along the length of the pipe. By dynamically adjusting the gauge length 332, the processing unit 103 may be able to reduce the signal-to-noise ratio in the detected backscattering signal and accurately identify the location of the leak along the pipe 1. This allows the processing unit 103 to classify any anomalous events or the inconclusive events as a particular event (e.g., a leak 2), as will be described with reference to Figures 5b, 6a and 6b.

[0113] Turning now to Figure 4 there is shown a schematic illustration of an example of a backscattering signal 400 detected at the light detector module 110. In this figure, the x-axis 404 represents a return signal wavelength and the y-axis represents a return signal intensity.

[0114] The backscattering signal 400 comprises a first peak 404 due to Rayleigh scattering of the light pulses, a second and third peaks 406 due to Brillouin scattering of the light pulses and fourth and fifth peaks 408 due to Raman scattering of the light pulses. The first peak 404 may be monitored in the DAS mode (by the Rayleigh spectrometer), the second and third peaks 406 may be monitored in the DSS and DTS modes (by Brillouin spectrometer) and the fourth and fifth peaks 408 may be monitored in the DTS mode (by the Raman spectrometer). As previously mentioned, in some embodiments, a single light detector module 110 (e.g., a single multimode spectrometer) may be configured to monitor all of the peaks 404, 406 and 408 in the backscattering signal 400.

[0115] Turning now to Figures 5 a and 5b, there are shown flow diagrams 500a and 500b of a method for monitoring the fluid pipe 1 in accordance with the present invention.

[0116] Turning specifically to flow diagram 500a. At block 502, at least one sensing fibre 10 is provided within the pipe 1. For example, the sensing fibre 10 may be inserted into the fluid pipe 1 and subsequently mounted to an inside surface of the fluid pipe 1.

[0117] At block 504, the light emitter 101 introduces light pulses with particular characteristics into the fibre 10. The particular characteristics of the light pulses include an initial: pulse width 344, pulse repetition frequency 342 and gauge length 332.

[0118] At block 506, the light detector module 110 operating in the DAS mode detects backscattering of the light pulses from the sensing fibre 10. In this embodiment the light detector module 110 may comprise a Rayleigh spectrometer 112 or a multi-mode spectrometer capable of operating in each of DAS, DTS and DSS modes.

[0119] At block 508, the light detector module 110, operating in the DAS mode, is configured to output a DAS backscattering signal in response to the detected backscattering of the light pulses. The DAS backscattering signal includes the first peak 404 generated by Rayleigh scattering of the light pulses.

[0120] Turning specifically to flow diagram 500b. The subsequent steps outlined in the flow diagram 500b are carried out by the processing unit 103.

[0121] At block 510, the processing unit 103, processes the backscattering signal to identify backscattering signal features characteristic of a particular event. In one example, the processing unit 103 processes the signal 400 and identifies that the peak 404 is indicative of a particular event (e.g., the leak 2). More specifically, the processing unit 103 can be configured to compare the determined vibration frequency and amplitude of the DAS detection signal 400 against expected vibration profiles of particular events. This can enable detection or exclusion of particular sources of vibration. In one example, this could involve matching the DAS backscattering signal 400 (and specifically the peak 404) to preset orifice noise profiles, flow noise profiles and / or negative pressure waves indicative of leaks.

[0122] At block 512, the processing unit 103 determines whether an anomalous event has been detected. The method for detecting an anomalous event will be described in more detail with reference to Figure 6a.

[0123] If the anomalous event has been detected the method proceeds to block 516. If the anomalous event has not been detected the method proceeds back to block 510.

[0124] At block 514, the processing unit 103 determines whether an inconclusive event has been detected. The method for detecting an inconclusive event will be described in more detail with reference to Figure 6b.

[0125] If the inconclusive event has been detected the method proceeds to block 516. If the inconclusive event has not been detected the method proceeds back to block 510.

[0126] At block 516, the processing unit 103 causes the light emitter 101 to vary the characteristics of the emitted light pulses. Varying the characteristics of the emitted light pulse 330 may comprise reducing or increasing: the pulse width 344, the pulse repetition frequency 342 and / or the gauge length 332, as described with reference to Figures 3c and 3d.

[0127] After varying the characteristics of the emitted light pulse 330, the processing unit 103 is configured to process the backscattering signal detected by light detector module 110 at a later point in time to try and classify the anomalous event or the inconclusive event as the particular event.

[0128] When an anomalous or inconclusive event is detected, adjusting the characteristics of the light pulses can refine the resolution and sensitivity of the measurements. More specifically, adjusting the characteristics of the light pulses can either increase or reduce the effective sensitivity of the apparatus 100, thereby reducing the impact of any prevailing background noise (like vibrations 22) on the backscattering signal detected by light detector module 110. This may allow the processing unit 103 to correctly identify the anomalous or inconclusive event as a particular event.

[0129] Upon not being able to classify the anomalous event or the inconclusive event after varying the characteristics of the emitted light pulse 330, the method may proceed to block 518. If the processing unit 103 successfully classifies the anomalous event or the inconclusive event after switching the sensing modes, the method may not proceed to block 518, but instead return to block 510. As such, proceeding to step 518 is optional.

[0130] At block 518, the processing unit 103 causes the light detector module 110 to switch to a different sensing mode. More specifically, the processing unit 103 causes the light detector module 110 to switch to the DTS or DSS sensing mode. This step may involve switching from detecting the backscattering signal using the Rayleigh spectrometer 112 to detecting the backscattering signal using the Raman spectrometer 114 or Brillouin Spectrometer 113.

[0131] Switching the detection modes offers significant advantages in increasing the confidence level of detecting and classifying particular events, such as leaks. By switching between multiple sensing modes (upon detecting an anomalous event or an inconclusive event), the system can gather diverse types of data, each providing unique insights into the physical state of the pipe. This multimodal approach allows for crossverification of detected anomalies or inconclusive events, enhancing the accuracy and reliability of event identification

[0132] After switching the sensing modes, the processing unit 103 is configured to process the backscattering signal detected by light detector module 110 at a later point in time to classify the anomalous event or the inconclusive event as the particular event.

[0133] Upon not being able to classify the anomalous event or the inconclusive event after switching the sensing modes, the method may proceed to block 520. If the processing unit 103 successfully classifies the anomalous event or the inconclusive event after switching the sensing modes, the method may not proceed to block 520, but instead return to block 510. As such, proceeding to step 520 is optional. At block 520, the processing unit 103 is configured to average the backscattering signal 400 over a period of time to output an averaged backscattering signal. Subsequently, the processing unit 103 is configured to use the averaged backscattering signal to classify the anomalous event or the inconclusive event as the particular event.

[0134] Averaging the backscattering signal over a period of time offers several advantages in accurately classifying an anomalous or inconclusive event. When initial and subsequent analyses fail to conclusively identify the event, continuous processing and averaging the signal helps to mitigate the effects of transient noise and random fluctuations. This approach enhances the signal-to-noise ratio, making persistent and consistent features of the signal more prominent while reducing the impact of sporadic or erratic disturbances.

[0135] Upon not being able to classify the anomalous event or the inconclusive event after averaging the backscattering signal 400, the method may proceed to block 522. If the processing unit 103 successfully classifies the anomalous event or the inconclusive event after averaging the backscattering signal 400, the method may not proceed to block 522, but instead return to block 510. As such, proceeding to step 522 is optional.

[0136] At block 522, the processing unit 103 is configured to employ a processing algorithm. The processing algorithm may be configured to extend the dynamic range of the backscattering signal.

[0137] Turning now to Figure 6a, which shows a flow diagram of a method 512 for detecting an anomalous event.

[0138] At block 600, the processing unit 103, determines typical backscattering signal features associated with the pipe 1 or a portion of the pipe 1. In some embodiments, the processing unit 103 may monitor the backscattering signal for a first period of time. Subsequently, the processing unit 103 may determine the typical backscattering signal features associated with the pipe 1 or a portion of the pipe 1 based on the backscattering signals which have been detected over the first period of time. At block 602, the processing unit 103, processing the backscattering signal at a point in time after the first period of time to identify a second group of backscattering signal features.

[0139] At block 604, the processing unit 103, compares the second group of backscattering signal features to the typical (or historical) backscattering signal features.

[0140] At block 606, the processing unit 103, determines that the difference between characteristics of the second group of backscattering signal features and the typical backscattering signal features is above a pre-defined threshold. For example, the processing unit may determine that the difference in amplitude and frequency of the second group of backscattering signal features and the typical backscattering signal features are above a pre-defined threshold.

[0141] In one example, backscattering signal features indicative of a large leak may be detected using the DAS mode, but no preceding buildup or warning signs may be observed by the apparatus. The large leak (in view of the lack of prior warning signs) may comprise an anomalous event. In some embodiments, the anomalous event may comprise any event that is not expected (or predicted by) the detected historical backscattering signals.

[0142] At block 608, the processing unit 103 determines that the second group of backscattering signal features are indicative of the anomalous event.

[0143] Advantageously, the process of detecting anomalous events by the processing unit involves a comprehensive approach aimed at distinguishing abnormal signal features from typical ones associated with the pipe or its operation. By initially monitoring the backscattering signal over a predefined period, the apparatus may establish a baseline of typical signal characteristics. Subsequently, processing the signal at a later time, may allow for the identification of deviations or anomalies, which may indicate potential issues or events requiring further investigation. Even if the detected signal features surpass the high confidence level, further scrutiny may be warranted to accurately identify the specific event. Turning now to Figure 6b, which shows a flow diagram of a method 514 for detecting an inconclusive event.

[0144] At block 610, the processing unit 103 assigns a confidence level to each of the identified backscattering signal features within the backscattering signal 400. Assigning a confidence level to each of the identified backscattering signal features within the backscattering signal 400 can be achieved through various analytical techniques. Initially, the backscattering signal may be processed to extract key features such as intensity, frequency, phase shift, and temporal patterns of any backscattering signal features (404, 406, 408) within the backscattering signal (400). Each of these features is then compared against a database of known signal characteristics associated with typical events, such as leaks, in-pipe and near pipe events, structural deformations, and normal operational states. Statistical methods, machine learning algorithms, or pattern recognition techniques can be employed to evaluate the similarity between the observed features and the reference characteristics.

[0145] For each identified backscattering feature (404, 406, 408), a confidence level is calculated based on the degree of match or correlation with known patterns. This confidence level represents the likelihood that a particular feature (404, 406, 408) is indicative of a specific event. For instance, if a feature closely matches the profile of a known leak, it is assigned a high confidence level. Conversely, if the feature deviates significantly from known event profiles, it is assigned a low confidence level. Features falling within an intermediate range of similarity are assigned intermediate confidence levels, indicating uncertainty and the potential need for further analysis.

[0146] At block 612, the processing unit 103, determines that the confidence level of a first backscattering signal feature (for example the peak 404) or a first group of backscattering signal features (for example peaks 406 or 408) lie in an intermediate confidence range. The backscattering signal features may lie below the high confidence level (above which the events are conclusively identified or classified) due to background noise in the signal (e.g., vibrations 22) or because the leak 2 is small.

[0147] At block 614, the processing unit 103, determines that the first backscattering signal feature (404) or the first group of backscattering signal features (406, 408) are indicative of an inconclusive event. It will be understood that the invention is not limited to the examples and embodiments above-described and various modifications and improvements can be made without departing from the concepts described herein. Except where mutually exclusive, any of the features may be employed separately or in combination with any other features and the disclosure extends to and includes all combinations and subcombinations of one or more features described herein. In particular, the sequence of operations shown in Figures 5a, 5b, 6a and 6b are merely exemplary. Any of the operations shown in the method 500a, 500b, 512 and 514 may be performed in a different order that achieves substantially the same result.

Claims

CLAIMS1. An apparatus for monitoring a fluid pipe, the apparatus comprising: at least one sensing fibre provided within the pipe; a light emitter for introducing light pulses with particular characteristics into the fibre; a light detector module configured to detect backscattering of the light pulses from the sensing fibre in multiple different sensing modes and output a backscattering signal in response thereto; a processing unit configured to process the backscattering signal to identify backscattering signal features characteristic of a particular event, wherein upon detecting an anomalous event or an inconclusive event the processing unit is configured to control the light detector module to switch between the sensing modes and / or to control the light emitter to vary the characteristics of the emitted light pulses.

2. An apparatus according to claim 1, wherein the particular event may comprise one or more of: a leak in the pipe; third party intrusion events; a change in external conditions proximal to the pipe; the pipe bursting; a structural deformation of the pipe; a change in fluid flow within the pipe; a change in fluid pressure within the pipe; a blockage in the pipe; and / or valve and pump operation.

3. An apparatus according to claim 2, wherein the processing unit is configured to process the backscattering signal to distinguish between different types of leaks in the pipe, including longitudinal cracks, and circumferential breaks.

4. An apparatus according to any one of the preceding claims, wherein after switching between the sensing modes and / or varying the characteristics of the emitted light pulses, the processing unit is configured to: process the backscattering signal detected by light detector module at a second point in time to classify the anomalous event or the inconclusive event as the particular event.

5. An apparatus according to claim 4, wherein upon not being able to classify the anomalous event or the inconclusive event at the second point in time, the processing unit is configured to: continue processing the backscattering signal until a third point in time; average the backscattering signal over a period of time between the second point in time and the third point in time to output an averaged backscattering signal; use the averaged backscattering signal to classify the anomalous event or the inconclusive event as the particular event.

6. An apparatus according to claim 4 as dependent on claim 2 or claim 3, wherein the processing unit is configured to generate an alarm in response to classifying the anomalous event or the inconclusive event as a leak in the pipe, the pipe bursting, a structural deformation of the pipe or a blockage in the pipe.

7. An apparatus according to any one of the preceding claims, wherein the multiple different sensing modes comprise at least: a distributed acoustic sensing (DAS)mode, a distributed strain sensing (DSS) mode and a distributed temperature sensing (DTS) mode.

8. An apparatus according to claim 7, wherein: the light detector module is configured to detect the backsc altering of the light pulses using the DAS mode and output a DAS backscattering signal in response thereto; and the processing unit is configured to process the DAS backscattering signal to determine an acoustic vibration experienced by particular points on the sensing fibre, and hence particular locations along the pipe, and to compare the determined acoustic vibration against expected acoustic profiles of particular events to identify the particular event.

9. An apparatus according to claim 7 or claim 8, wherein: the light detector module is configured to detect the backscattering of the light pulses using the DSS mode and output a DSS backscattering signal in response thereto; and the processing unit is configured to process the DSS backscattering signal to determine strain induced variations experienced by particular points on the sensing fibre, and hence particular locations along the pipe, and to compare the determined strain induced variations against expected strain profiles of particular events to identify the particular event.

10. An apparatus according any one of claims 7 to 9, wherein: the light detector module is configured to detect the backscattering of the light pulses using the DTS mode and output a DTS backscattering signal in response thereto; and the processing unit is configured to process the DTS backscattering signal to determine temperature variations experienced by particular points on the sensing fibre, and hence particular locations along the pipe, and to compare the determined temperature variations against expected temperature profiles of particular events to identify the particular event.

11. An apparatus according to any one of claims 7 to 10, wherein the light detector module is configured to initially detect backscattering of the light pulses from the sensing fibre using the DAS mode, and wherein the processing unit is configured to control the light detector module to switch to the DSS mode or the DTS mode upon detecting the anomalous event or the inconclusive event.

12. An apparatus according to any one of claims 7 to 11, wherein upon detecting the anomalous event or the inconclusive event the processing unit is configured to: vary the characteristics of the emitted light pulses; determine whether varying the characteristics of the emitted light pulses leads to a successful classification of the anomalous event or the inconclusive event as the particular event; and upon determining that varying the characteristics of the emitted light pulses does not lead to a successful classification of the anomalous event or the inconclusive event, causing the light detector module to switch between the sensing modes.

13. An apparatus according to any one of the preceding claims, wherein varying the characteristics of the emitted light pulses comprises reducing or increasing one or more of: a pulse width; a pulse repetition frequency; and / or a gauge length.

14. An apparatus according to any one of the preceding claims, wherein detecting the inconclusive event comprises: assigning a confidence level to each of the identified backscattering signal features within the backscattering signal;determining that the confidence level of a first backscattering signal feature or a first group of backscattering signal features lies in an intermediate confidence range; and determining that the first backscattering signal feature or the first group of backscattering signal features are indicative of an inconclusive event.

15. An apparatus according to any one of the preceding claims, wherein detecting the anomalous event comprises: monitoring the backscattering signal for a first period of time; determining typical backscattering signal features associated with the pipe or a portion of the pipe based on the monitoring; processing the backscattering signal at a point in time after the first period of time to identify a second group of backscattering signal features; comparing the second group of backscattering signal features to the typical backscattering signal features; determining that that the second group of backscattering signal features differs from the typical backscattering signal features; and determining that the second group of backscattering signal features are indicative of the anomalous event.

16. An apparatus according to claim 15, wherein determining that that the second group of backscattering signal features differs from the typical backscattering signal features comprises: determining that the difference between characteristics of the second group of backscattering signal features and the typical backscattering signal features is above a pre-defined threshold.

17. An apparatus according to any one of the preceding claims, wherein the sensing fibre comprises a single fibre or a dedicated fibre within a bundle of fibres.

18. An apparatus according to any one of the preceding claims, wherein a barrier is provided between the sensing fibre and the fluid within the pipe.

19. An apparatus according to any one of the preceding claims, wherein the sensing fibre is fixedly connected to an inside surface of the pipe.

20. A method for monitoring a fluid pipe, the method comprising the steps of: providing at least one sensing fibre within the pipe; introducing light pulses with particular characteristics into the fibre using a light emitter; detecting backscattering of the light pulses from the sensing fibre using a light detector module operating in a first sensing mode; outputting, by the light detector module, a backscattering signal in response to the detected backscattering of the light pulses; processing, by a processing unit, the backscattering signal to identify backscattering signal features characteristic of a particular event; detecting, by the processing unit, an anomalous event or an inconclusive event; and in repose to detecting the anomalous event or the inconclusive event causing the light detector module to switch to a different sensing mode and / or causing the light emitter to vary the characteristics of the emitted light pulses.

21. A method according to claim 20, wherein the first sensing mode comprises a distributed acoustic sensing (DAS) mode and the different sensing mode comprises either a distributed strain sensing (DSS) mode or a distributed temperature sensing (DTS) mode.

22. A method according to claim 20 or claim 21 , wherein varying the characteristics of the emitted light pulses comprises reducing or increasing one or more of: a pulse width, a pulse repetition frequency; and / or a gauge length.

23. A method according to any one of claims 20 to 22, wherein after switching to the different sensing mode and / or varying the characteristics of the emitted light pulses, the method comprises: processing, by the processing unit, the backscattering signal detected by light detector module at a second point in time; and classifying the anomalous event or the inconclusive event as a particular event.

24. A method according to claim 23, the method further comprising: generating, by the processing unit, an alarm in response to classifying the anomalous event or the inconclusive event as a leak in the pipe, the pipe bursting, a structural deformation of the pipe or a blockage in the pipe.

25. A pipe network comprising one or more pipes monitored using the apparatus of any one of claims 1 to 19 and / or the method of any one of claims 20 to 24.

Images