Method and system of leak detection

A system for detecting and locating a fluid leakage in a fluid pipeline network, the system comprising:

• a plurality of water meters installed in an area of the pipeline network,
• a sensor connected to each of the plurality of water meters, each sensor being configured to output data relating to an acoustic or vibration measurement associated with a part of the pipeline network proximate said sensor,
• wherein at least one of the water meter and the sensor is configured to record the acoustic or vibration measurement output data, transform the output data from a time domain to a frequency domain, and send the output data to a remote server,
• wherein the transformed output data in the frequency domain is separated into discrete frequency bins each containing a range of frequency values, and a peak magnitude value of each frequency bin is recorded, and
• wherein the remote server comprises a processor configured to:
• determine a leakage condition by comparing, within a predetermined frequency range, the peak magnitude values of the transformed output data in the frequency domain with predetermined threshold values for corresponding frequency bins,
• compute, for each sensor of the plurality of sensors, a sum value of the peak magnitude values within a predetermined frequency range of the transformed output data in the frequency domain and associating said sum value of the peak magnitude values with a corresponding sensor of said plurality of sensors,
• locate a source of the fluid leakage by identifying at least two sensors of the plurality of sensors in the area of the pipeline network that-is are closest to the source of the fluid leakage, wherein the at least two sensors closest to the source of the fluid leakage are identified based on comparing said sum values of the peak magnitude values across the plurality of sensors in the area of the pipeline network over a period of time, and
• estimate a location of the source of the fluid leakage based on (1) a relative geographical pipeline distance between the identified sensors, (2) configuration of pipelines between the identified sensors, and (3) the identified sensors' respective sum values.

A remote server when used for detecting and locating a fluid leakage in a fluid pipeline network, the server comprising a processor configured to:

• receive output data, from at least one of a plurality of water meters and sensors in an area of the pipeline network, relating to acoustic or vibration measurements as measured by sensors associated with a part of the pipeline network proximate said sensors,
• transform any time domain output data to data in a frequency domain, wherein the transformed output data in the frequency domain is separated into discrete frequency bins each containing a range of frequency values, and a peak magnitude value each frequency bin is recorded,
• determine a leakage condition by comparing, within a predetermined frequency range, the peak magnitude values of the transformed output data in the frequency domain with a predetermined threshold value for a corresponding frequency bins,
• compute, for each sensor of the plurality of sensors, a sum value of the peak magnitude values within a predetermined frequency range of the transformed output data in the frequency domain and associating said sum value of the peak magnitude values with a corresponding sensor,
• locate a source of the fluid leakage by identifying at least two sensors in the area of the pipeline network that is are closest to the source of the fluid leakage, wherein the at least two sensors closest to the source of the fluid leakage are identified based on comparing said sum values of the peak magnitude values across the plurality of sensors in the area of the pipeline network over a period of time, and
• estimate a location of the source of the fluid leakage based on (1) a relative geographical pipeline distance between the identified sensors, (2) configuration of pipelines between the identified sensors, and (3) the identified sensors' respective sum values.