Abstract

Leakage of water pipelines will significantly endanger the safety operation and service performance of the pipelines. Based on the vibration of pressurized water pipelines deriving from leakage, the BA-FH3200 fiber optic hydrophone (FOH) leakage detection long-term detection system was adopted in prototype tests. The vibration-based real-time leakage monitoring method of the pressurized water pipeline was studied. During the test, the leakage was simulated by opening a spherical valve in the middle of the pipe, and an FOH was placed right above the pipe wall to detect the vibration signal along the pipe. The FOH analysis software was used to monitor the pipeline operation status in real time and acquire data. Then, the data were processed by a self-developed post-processing program, and the parameters were optimized through back-calculation. The test results reveal that the leakage positioning error lay between ±0.07 m, and real-time monitoring (i.e., early warning alarm and leakage positioning) of the FOH for the pressurized water pipeline was feasible.

INTRODUCTION

Pressurized water pipelines exhibit various advantages, e.g., smooth and reliable transportation, no susceptibility to pollution, low transportation costs, saving of land, and convenient management and maintenance, therefore they are widely utilized in waterworks. As closed systems, transmission mains (TMs) may experience large overpressures due to water hammer, while water distribution networks (WDNs) can dissipate the water pressure based on open systems. Thus, TMs and WDNs will exhibit different responses during operation. The material of buried pipes may include cast iron, ductile iron, asbestos cement, and polyvinyl chloride. Due to the aging of pipe systems, material loss and artificial destruction, etc., leakage issues in water pipelines occur frequently (Zhang & Fu 2002; Zhang et al. 2006; Zhang & Zhang 2008) and consequently, hydraulic capacity, safety operation and service performance of pipelines is seriously endangered (Al-Barqawi & Zayed 2006). Severe losses (e.g. water loss, water contamination and energy consumption) (Colombo & Karney 2002) may be induced. Thus, it has been an urgent issue for current waterworks to strengthen the real-time monitoring of the key pipeline sections of built pipe networks. It will refine the acquisition of early warning alarm information and performance in emergency disposal.

Currently, pipeline leakage detection generally ranges from physical inspection to advanced satellite imaging. Regarding the type of detection methods, this can be divided into two groups, i.e., hardware-based and software-based (Datta & Sarkar 2016). Among the hardware-based approaches, special sensing equipment is commonly used to detect pipeline leakage, as follows. (1) Based on piezoelectric acoustic emission (AE) sensors. The new method of AE source positioning based on AE signal correlation function (Grabec 1978) can effectively promote the positioning ability of the whole system. Combined with simple signal processing (Wu et al. 2018), AE technology is also feasible for identifying industrial underground pipeline leakage (Lim 2015). The reference standard for evaluating AE sensor equipment in pipeline leakage detection has been gradually established and refined (Miller et al. 1999). AE technology can be applied to the detection and positioning of pressurized pipeline defects during normal operation, while it is insensitive to the geometrical structure of the tested parts. (2) Based on the leak detector. Since the principle of the leak detector was put forward (Guest 1972), various studies have been conducted on detection methods. Different detectors have been developed, e.g., the acoustic valve leak detector (AVLD) (Dimmick et al. 1979) that detects fluid leakage, the leak detector that can achieve automatic leakage detection in the pipeline (Chatzigeorgiou et al. 2015a) and a new pipeline automatic leakage sensing system (the MIT leak detector) (Chatzigeorgiou et al. 2015b). (3) Based on ground-penetrating radar (GPR). GPR is increasingly employed as an effective tool for detecting pipeline leakage (Stampolidis et al. 2003). It is feasible to apply GPR to the leakage detection of buried water pipes through electromagnetic simulation (Nakhkash & Mahmood-Zadeh 2004). The laboratory GPR scaling model has also been established to evaluate the parameters involved in detecting water pipe leakage (Hyun et al. 2007). Furthermore, based on laboratory experiments, the disturbance mode of GPR images has been studied (Lai et al. 2016). Through the study by Amran et al. (2018), the favorable performance of GPR in detecting underground pipelines and locating leakage has been verified. (4) Based on fiber optic sensors. When a gas or liquid pipeline leaks, a wideband acoustic signal will be generated, and then the resulting signal will induce variation in the optical phase signal of the fiber fixed on the surface of the pipe (Huang et al. 2007). In terms of leakage in fluid-filled high voltage distribution lines, a distributed fiber optic acoustic sensor technology has been proposed for measurement and positioning (Kurmer et al. 1992). It has also been revealed that distributed optical fiber is feasible for continuously detecting leakage in a buried pressurized water pipeline (Jin et al. 2016). The effectiveness of fiber optic pressure sensors in transient pipeline leakage detection has also been verified (Gong et al. 2018).

Software-based detection methods rely on a wide variety of software programs, as follows. (1) Based on negative pressure wave (NPW). Based on the partial reflection of transient pressure waves from the leak, the location and discharge behaviour can be identified. Using a physical pipeline model, the variation of the NPW and its attenuation regularity along the pipeline have been deduced (Brunone 1999; Ge et al. 2008). A proposed small noise reduction method based on empirical mode decomposition (SNR-EMD) can reduce the noise in the pipeline pressure signal, and combined with SNR-EMD, the NPW can be applied to pipeline leakage positioning (Lu et al. 2016). Considering the influence of rubber washers on NPW, a numerical method for calculating the propagation velocity of a pressure wave in a liquid pipeline has also been proposed (Chen et al. 2018), and it can increase the leakage positioning accuracy of the assembled pressurized liquid pipeline. The obvious pressure variation at both ends of the pipe may be easily detected with the NPW-based method, while the pressure variation caused by small or slow leaks cannot be detected (Sun & Chang 2014). (2) Based on support vector machine (SVM). Combined with SVM, a pipeline leakage detection and early warning system has been proposed (Qu et al. 2010). With the addition of particle swarm optimization (PSO) theory, SVM was also able to detect the leakage position of a pipeline (Ni et al. 2013). In the PSO-SVM pipeline leakage diagnosis model, the PSO algorithm has also been applied to parameter optimization selection (Wang et al. 2018). In terms of leakage detection in a liquid pipeline, a leakage detection method based on mobile window least squares support vector machine (MWLS-SVM) has been proposed (Li et al. 2018). (3) Based on genetic algorithm (GA) and inverse transient analysis approach. In terms of inverse transient analysis, based on transient analysis, time-lagged calculations, inverse calculations, and event detection, the calibration and operational safety of pipelines can be continuously achieved, and especially, the inverse calculation is the most useful part and can calibrate while determining leaks or unauthorized use (Liggett & Chen 1994). Simpson et al. (1993) adopted GA to optimize the pipeline network and analyze the GA parameters. Combined with the inverse transient method (Liggett & Chen 1994), GA technology could optimize the pipe network (Simpson et al. 1994) and could be used to detect the leakage and friction coefficient in the water distribution system (Vítkovský et al. 2000). For transient-based techniques (TTBTS), it is often challenging to access and generate transients. The valves may be installed somewhere due to (e.g.) difficult pressure measurements or slow valve response. The pressure surges induced by the maneuver are not determined with sufficient precision in many valves. Thus, it is beneficial to find out how to introduce transients instead of valve control (Brunone et al. 2008). (4) Based on artificial neural network (ANN). ANN can be used to analyze the hydraulic parameter sensor data in a water distribution system (Mounce & Machell 2006). Incorporating the approximate entropy, an ANN leakage detection method can distinguish the leakage signal from the non-leakage signal (Yang et al. 2010). ANN can also estimate the pressure distribution in the water distribution system, and then, the optimum monitoring node can be determined by an entropy-based ANN method (Ridolfi et al. 2014).

Through analysis and comparison, defects in the current detection methods are also revealed. Pipeline leakage detection based on a piezoelectric AE sensor is difficult to implement under complex conditions. GPR is susceptible to disturbance by metal media around pipelines, and detection will involve high costs. Manhole dropping equipment needs to be provided for detection by leak detectors. The NPW-based detection method is not suitable for detecting in short-distance pipeline. A large amount of data needs to be acquired with the SVM-based detection method. For TTBTS, it is often challenging to access and generate transients. With regards to ANN, the detection process is too complicated. Overall, for leakage detection in a water pipeline, the current hardware- and software-based methods may still exhibit limitations and applicable conditions.

In this study, the long-term leakage detection system based on the BA-FH3200 fiber optic hydrophone (FOH) contains the FOH, FOH analyzer, a laptop, FOH analysis software and signal post-processing program. Based on the FOHs, the pipeline was monitored; the FOH analyzer would record the data once started during monitoring; through the self-developed post-processing program, the signal would be analyzed to obtain the time delay; based on the parameter optimization and error calculation, the leak positions would be determined. FOHs possess the advantages of high pressure-sensitivity, environmental adaptability and reliability, etc., and the specific parameters are shown in Table 1. The FOH analyzer is employed to demodulate the signal of the interferometric FOH based on the 3 × 3 demodulation principle. Its advantages include high precision, large dynamic range, low background noise and high demodulation consistency, as demonstrated in Table 2. Combined with the advantages of both to develop a systematic detection method, it may be more efficient to monitor water pipeline leakage.

Table 1

Technical indicators of fiber optic hydrophone

No.IndicatorDataUnit
Acoustic pressure sensitivity −133 dB rad/Hz 
Frequency response range 10–3000 Hz 
Sensitivity range in frequency band ≤ ±1.5 dB 
Dimensions Probe: Φ36 × 124 mm 
Operating temperature range 5–75 °C 
Water pressure resistance 10 MPa 
No.IndicatorDataUnit
Acoustic pressure sensitivity −133 dB rad/Hz 
Frequency response range 10–3000 Hz 
Sensitivity range in frequency band ≤ ±1.5 dB 
Dimensions Probe: Φ36 × 124 mm 
Operating temperature range 5–75 °C 
Water pressure resistance 10 MPa 
Table 2

Technical indicators of fiber optic hydrophone analyzer

No.IndicatorDataUnit
Sampling frequency 32 kHz 
Working wavelength 1550 ± 10 nm 
System self-noise −100 dB 
Optical interface FC/APC – 
Network interface TCP/IP – 
Storage mode Computer memory – 
No.IndicatorDataUnit
Sampling frequency 32 kHz 
Working wavelength 1550 ± 10 nm 
System self-noise −100 dB 
Optical interface FC/APC – 
Network interface TCP/IP – 
Storage mode Computer memory – 

FUNDAMENTAL PRINCIPLES

Fiber optic hydrophones

BA-FH3200 FOH is a commercial product developed by Shanghai Baian Sensing Technology based on Michelson fiber interferometer technology. FOH is a new generation of underwater acoustic transducer technology, and its working principle is as follows: through the optical fiber modulation technology, the sound or vibration signal is converted to optical signal phase information; then, through the optical fiber, the phase information is transmitted to the signal acquisition end machine; then, it is restored to the sound or vibration signal through optical phase demodulation. A schematic diagram of the principle of the FOH based on the Michelson fiber interferometer can be seen in Chu (2017). The laser signal is emitted by the laser; next, through a 3 dB fiber coupler, it is divided in two ways, i.e., one way as the sensing arm accepts the modulation of the sound waves and the other way as the reference arm provides the reference phase; then, the two beams of waves will return to the fiber coupler after being reflected by the back-end reflector, and further the interference occurs; finally, the interference signal is converted to an electrical signal by the detector, and the information on the acoustic signals can be acquired by signal processing (Zhang & Ni 2004).

The optical fiber is wound around the elastic body with air backing to form an acoustic sensitive area; when the underwater acoustic signal acts on the elastic body, micro-strain will be caused to the elastic body, and further to the single-mode optical fiber wound around the elastic body; the strain in the optical fiber will modulate the phase of the propagating lightwave in the fiber. Therefore, the size of the underwater acoustic signal can be obtained by demodulating the phase change of the optical wave.

The basic relationship of FOHs is expressed as: 
formula
(1)
where I is the output intensity of the fiber-optic Michelson interferometer, A is the DC component, B is the amplitude of the AC component caused by the laser interference, and ΦS is a phase shift of the light that can be modulated by the sound pressure, ΦS = 2πnL/λ, n is the refractive index of the fiber core, L is twice the length of the sensitive fiber, λ is the working wavelength of the laser, and Φ0 is the initial phase difference between the signal arm and the reference arm of the interferometer. The operation process of the sound-pressure fiber optic water acoustic sensor is to demodulate the change of the ΦS from the change of the output light intensity I of the fiber Michelson interferometer, and then the magnitude of the sound pressure P causing the change of S is determined.

Signal processing

When a water pipeline leaks, the leak-induced sound waves will propagate along both sides of the pipeline (Zhang & Ni 2004). Due to the distance difference, the sensors installed on both sides will receive signals with a time difference, i.e., the time delay. The delay value can be estimated with the correlation signal processing method. In terms of this issue, Knapp & Carter (1976) proposed a generalized correlation function method; based on the existing theory and pre-filter, the estimated delay value could be refined. The relationship between the setting of the pre-filter and the delay estimators was investigated.

The delay value is determined by adopting the cross-correlation function method (Zhang et al. 2003). The principle is to analyze the two noise signals collected by the sensors on both sides of the leakage point by the cross-correlation function, and the value corresponding to the maximum point in the cross-correlation function is the delay value. This method can realize the estimation of time difference, and the principle of the correlation function method is illustrated in Figure 1 (Knapp & Carter 1976).

Figure 1

Schematic diagram of the correlation function method.

Figure 1

Schematic diagram of the correlation function method.

In Figure 1, H1 and H2 are introduced as prefilters. Through the relationship between the Fourier transform and power spectral density, it can be obtained as: 
formula
(2)
The mutual power spectral density of signal x1 and x2 after H1 and H2 filtering is: 
formula
(3)
Thus, the broad cross-correlation result of x1 and x2 is: 
formula
(4)
where 
formula
(5)

TEST SET-UP

Test preparation

The test site is demonstrated in Figure 2(a)–2(f). In this test, DN300 ductile cast iron pipe was adopted with inner diameter of 300 mm, thickness of 5 mm, and elastic modulus of 140–154 GPa. Seventeen of the DN300 ductile iron pipes were used with a length of 6 m each. These pipes were connected in sequence from right to left (i.e., No. 1, 2, …, 17). In order to eliminate the effect of water inlet and outlet on the test results at both ends of the pipe, the FOHs were not mounted on the two pipes at the beginning and end. Right above the pipeline, FOHs (No. 1, 2, 3, 4) were installed. No. 1 and 2 FOHs were lying on the No. 3 and 15 pipes respectively, and No. 3 and 4 FOHs lay right above the leakage point and at a distance of 10 m from the leakage point respectively. Between the No. 9 and 10 pipes, a reducer tee was installed to simulate the leakage. The TCP/IP network interface on the FOH analyzer panel was connected to the laptop. The input and output interfaces connected FOH No. 1, 2, 3 and 4.

Figure 2

Layout of test site and equipment: (a) pipeline arrangement, (b) spherical valve, (c) inlet of pipeline with pressure gauge, (d) data acquisition system, (e) rear side of FOH analyzer and (f) fiber optic hydrophone.

Figure 2

Layout of test site and equipment: (a) pipeline arrangement, (b) spherical valve, (c) inlet of pipeline with pressure gauge, (d) data acquisition system, (e) rear side of FOH analyzer and (f) fiber optic hydrophone.

Test process

The schematic diagram of the test site layout is illustrated in Figure 3. The test is started by opening the discharge valve at the end of the pipe and turning on the pressurized water pump to fill the pipe with water; adjusting and controlling the opening degree of the discharge valve at the end; waiting until the indications of the three pressure gauges are steady, and fixing the opening degree of the discharge valve at the end of the pipe. Then, four FOHs are mounted right above the pipe wall. When the pipeline reaches the steady state without vibration, data acquisition will be conducted. Next, the pipeline will be in abnormal operation with simulated leakage, and the relevant leak detection test will be performed. In addition, the data will be automatically acquired. The leakage range of the pipe will be simulated by controlling the opening of the spherical valve in the middle of the pipe.

Figure 3

Schematic diagram of the test site layout (unit in m, except ø in mm).

Figure 3

Schematic diagram of the test site layout (unit in m, except ø in mm).

Data acquisition

Before data acquisition, the FOH analyzer starts; the IP address and command code will be input; and the FOH analysis software is run on the laptop. The specific interface is shown in Figure 4.

Figure 4

Main interface of the fiber optic hydrophone analysis software.

Figure 4

Main interface of the fiber optic hydrophone analysis software.

Next, delay calibration, calibration, parameter dispatch, demodulation, contrast, etc., are performed. Then, the wave data are saved as a DAT file every minute.

The normal demodulation interface will provide an early warning alarm if there exists any leakage in the pipeline. No. 1, 2, 3 and 4 images in the demodulation interface demonstrate the operation of No. 1, 2, 3 and 4 FOHs respectively during the monitoring of the pipelines. According to the normal operation threshold of the pipe network, the warning threshold and the alarm threshold are set. At normal operation, the amplitude, frequency and threshold are within a certain range; when the pipeline state changes, the corresponding value changes accordingly. When the pipeline leaks, the absolute values of the corresponding signal increase, and then the lower alarm indicator lights up, as shown in Figure 5.

Figure 5

Demodulation interface of the fiber optic hydrophone analysis software.

Figure 5

Demodulation interface of the fiber optic hydrophone analysis software.

POST-PROCESSING PROGRAM

Visual Basic (VB) from Microsoft Corporation possesses the advantages of fast, efficient, and visual design. The VB-based self-developed post-processing program interface includes basic parameters, delay value estimation, leak location and parameter optimization, as shown in Figure 6.

Figure 6

Main interface of pipeline leakage monitoring.

Figure 6

Main interface of pipeline leakage monitoring.

Leakage positioning

As the water pipeline leaks, local vibration is induced at the leakage point by the pressure difference between the inside and outside of the pipeline, and then the whole pipeline generates vibration with different frequencies. The vibration propagates rapidly to both ends of the pipeline in the form of sound waves (Smith 2010). The FOHs are arranged at both ends of the water pipeline, and the leakage will be recognized once leakage occurs (Chu 2017). Further, the leakage positioning can be performed. The positioning principle is as illustrated in Figure 7.

Figure 7

Schematic diagram of sound wave leakage positioning.

Figure 7

Schematic diagram of sound wave leakage positioning.

The following relationships exist between the variables in the pipeline, where t1 and t2 are the time when the hydrophones detect leakage noise on both sides of the pipeline leakage point, assuming t1 > t2. 
formula
(6)
 
formula
(7)
 
formula
(8)
Therefore, the distance from the upstream point of the leakage point can be obtained, i.e., the calculation formula of the predetermined position is: 
formula
(9)
where L is the distance from the upstream point in m, S is the spacing between the hydrophones in m, V is leakage wave velocity in m/s, and Td is the time delay value in s. According to Equation (9), V and Td have significant influences on L, and therefore, the two key parameters should be given much attention.
In the study of Watters (1984), consideration was given to the propagation velocity of sound waves and the temperature, the inner diameter of the pipe, the thickness of the pipe wall, and the Young's modulus of the pipe. The formula for calculating the velocity of the leaking sound wave in the pipe is expressed as: 
formula
(10)
where K is the bulk modulus of water in N/m2, ρ is water density in kg/m3, D is the pipe's inner diameter in m, e is the pipe wall thickness in m, E is Young's modulus of elasticity of the pipe material in N/m2, and C is a dimensionless parameter describing the effect of pipe restraint on the speed of sound wave propagation. For the water supply pipe, C is taken as 1. The water hammer velocity of the DN500 (in mm) steel pipe in the hydraulic model established by Li & Zhao (2001) is 1172 m/s. Based on the calculation, the wave speed of DN300 ductile cast iron pipes is about 1016 m/s. According to the relationship between pressure wave speed and steel pipe diameter by Meniconi et al. (2015), the above speed value through calculation is suitable for the diameter.

Based on Equations (9) and (10), the distance of the leakage from the specific point can be determined. Then, it will be adopted as a reference for pipeline maintenance. As can be seen from Equation (10), the wave velocity will also vary with the material property of the pipeline. Therefore, positioning accuracy may depend on the material.

Time delay estimation

The delay value Td is calculated by the cross-correlation function. The cross-correlation function in signal analysis represents the degree of correlation between two time series and plays a significant role in practice. Due to the different transmission distances, the FOHs placed on the pipe wall will receive signals at different times, resulting in a time difference. Based on the digital graphics processing and signal processing functions of MATLAB from MathWorks, the original signal image and cross-correlation function image are generated from the previously saved DAT file and the delay value is achieved. After calling MATLAB, the image and time delay will be displayed, as shown in Figure 8.

Figure 8

Time delay estimation by calling MATLAB.

Figure 8

Time delay estimation by calling MATLAB.

Parameter optimization

The wave velocity factor C is a dimensionless parameter that describes the effect of pipe restraint on the propagation velocity of the sound wave. It is taken as 1 for the water supply pipe (Watters 1984). When the operating conditions change, this parameter may cause a large error. Therefore, the parameter needs to be optimized. According to the inverse algorithm, Equation (11) is obtained from Equations (9) and (10): 
formula
(11)

RESULTS AND DISCUSSION

After optimization, i.e., C= 1.1, the error can be effectively reduced, i.e., the error decreased from 0.442% to 0.171% with C from 1 to 1.1, and the specific values and test results are shown in Table 3.

Table 3

Comparison of leakage positioning before/after parameter optimization

Simulated leakage pointTime delay/sLeakage distance/m
Real distance/mError %/ ± m
BeforeAfterBeforeAfter
No. 1 & 2 hydrophones 0.0098 42.989 42.873 42.8 0.442% (0.189) 0.171% (0.073) 
No. 3 & 4 hydrophones 0.032 42.718 42.797 0.192% (−0.082) 0.007% (−0.003) 
Simulated leakage pointTime delay/sLeakage distance/m
Real distance/mError %/ ± m
BeforeAfterBeforeAfter
No. 1 & 2 hydrophones 0.0098 42.989 42.873 42.8 0.442% (0.189) 0.171% (0.073) 
No. 3 & 4 hydrophones 0.032 42.718 42.797 0.192% (−0.082) 0.007% (−0.003) 

From the test results, it is shown that the positioning error decreased slightly with the smaller distance.

The time delay values showed a downward trend with the distance of the hydrophones. This indicates that the distance could be an important factor in technique performance. After the parameter optimization, the error was reduced, i.e., 0.271% and 0.185% respectively for No. 1 and 2 and No. 3 and 4 hydrophones.

It is consistent with the fact that the wave may attenuate with the distance, and thus, the error slightly increased. In addition, it may also help to reveal the high precision of the adopted fiber optic hydrophones, and further research is still needed to better understand the results.

Through the post-processing analysis of the data collected during the warning alarm period, the long-term leak detection system based on the BA-FH3200 FOH can quickly and accurately detect the leakage location. The information will be provided for the management department and the corresponding response will be made according to the leak size.

CONCLUSIONS

In this study, DN300 ductile cast iron pipe was used as the test pipe, and a long-term leakage detection system based on the BA-FH3200 FOH was utilized to monitor the running condition of the pipeline in real time. Based on the model test, the results indicate the following:

  • (1)

    The BA-FH3200 FOH long-term leakage detection system could provide real-time warning/alarm during monitoring of a leakage in the water pipeline. It could detect the leakage location rapidly and accurately once leakage occurred.

  • (2)

    Based on MATLAB and VB, the self-developed post-processing program could achieve programmatic and visual data analysis processing and error correction. It may help promote the detection system to be more powerful in leakage analysis.

  • (3)

    According to the inverse calculation, the value of the wave velocity factor C was optimized. When C was set as 1.1, the error of leakage positioning could be controlled within the range of ±0.07 m.

  • (4)

    This paper investigated the feasibility of FOH for detecting pipeline leakage and may provide a new technical reference for leakage detection in a water pipeline.

There are still limitations exhibited with regard to this technique. In real practice, there may be multiple leaks in the pipeline at the same time. The sound waves will have influences on each other. Further research is still necessary to accurately identify the leakage location with multiple signal interference and determine the accuracy for different pipeline materials. More conditions will also need to be considered, e.g., water pressure, pipe diameter, leak size, sensor combination, sensor distance and changes in materials and branches in pipelines.

ACKNOWLEDGEMENTS

The authors would like to thank all the technicians from Water Conservancy and Transportation Infrastructure Safety Protection Henan Province Collaborative Innovation Center for the assistance in preparing the model tests.

REFERENCES

REFERENCES
Al-Barqawi
H.
Zayed
T.
2006
Condition rating model for underground infrastructure sustainable water mains
.
Journal of Performance of Constructed Facilities
20
(
2
),
126
135
.
Amran
T. S. T.
Ismail
M. P.
Ahmad
M. R.
Amin
M. S. M.
Ismail
M. A.
Sani
S.
Masenwat
N. A.
Basri
N. S. M.
2018
Monitoring underground water leakage pattern by ground penetrating radar (GPR) using 800 MHz antenna frequency
.
IOP Conference Series: Materials Science and Engineering
298
,
012002
.
Brunone
B.
1999
Transient test-based technique for leak detection in outfall pipes
.
Journal of Water Resources Planning and Management
125
(
5
),
302
306
.
Brunone
B.
Ferrante
M.
Meniconi
S.
2008
Portable pressure wave-maker for leak detection and pipe system characterization
.
Journal of American Water Works Association
100
(
4
),
108
116
.
Chatzigeorgiou
D.
Youcef-Toumi
K.
Ben-Mansour
R.
2015a
Design of a novel in-pipe reliable leak detector
.
IEEE/ASME Transactions on Mechatronics
20
(
2
),
824
833
.
Chatzigeorgiou
D.
Youcef-Toumi
K.
Ben-Mansour
R.
2015b
MIT leak detector: modeling and analysis toward leak-observability
.
IEEE/ASME Transactions on Mechatronics
20
(
5
),
2391
2402
.
Chen
Q.
Shen
G.
Jiang
J.
Diao
X.
Wang
Z.
Ni
L.
Dou
Z.
2018
Effect of rubber washers on leak location for assembled pressurized liquid pipeline based on negative pressure wave method
.
Process Safety and Environmental Protection
119
,
181
190
.
Chu
X.
2017
Seepage Monitoring Research on Long-Distance Water Transmission Pipeline Based on Noise Signal
.
Master's degree
,
Zhengzhou University, China
.
Colombo
A. F.
Karney
B. W.
2002
Energy and costs of leaky pipes: toward comprehensive picture
.
Journal of Water Resources Planning and Management
128
(
6
),
441
450
.
Datta
S.
Sarkar
S.
2016
A review on different pipeline fault detection methods
.
Journal of Loss Prevention in the Process Industries
41
,
97
106
.
Dimmick
J. G.
Nicholas
J. R.
Dickey
J. W.
Moore
P. M.
1979
Acoustical valve leak detector for fluid system maintenance
.
Naval Engineers Journal
91
(
2
),
71
83
.
Gong
J.
Png
G. M.
Arkwright
J. W.
Papageorgiou
A. W.
Cook
P. R.
Lambert
M. F.
Simpson
A. R.
Zecchin
A. C.
2018
In-pipe fibre optic pressure sensor array for hydraulic transient measurement with application to leak detection
.
Measurement
126
,
309
317
.
Guest
R. J.
1972
Leak detector and method. US3691819A, Google Patents
.
Huang
S.-C.
Lin
W.-W.
Tsai
M.-T.
Chen
M.-H.
2007
Fiber optic in-line distributed sensor for detection and localization of the pipeline leaks
.
Sensors and Actuators A: Physical
135
(
2
),
570
579
.
Hyun
S.-Y.
Jo
Y.-S.
Oh
H.-C.
Kim
S.-Y.
Kim
Y.-S.
2007
The laboratory scaled-down model of a ground-penetrating radar for leak detection of water pipes
.
Measurement Science and Technology
18
(
9
),
2791
2799
.
Jin
K.
Shang
F.
Lu
Z.
Zhu
X.
2016
Prototype test of distributed optical fiber continuous monitoring for buried pressure water conveyance pipeline
.
Water Resources and Hydropower Engineering
47
(
11
),
107
110
.
Knapp
C.
Carter
G.
1976
The generalized correlation method for estimation of time delay
.
IEEE Transactions on Acoustics, Speech, and Signal Processing
24
(
4
),
320
327
.
Kurmer
J. P.
Kingsley
S. A.
Laudo
J. S.
Krak
S. J.
1992
Distributed fiber optic acoustic sensor for leak detection
. In:
Distributed and Multiplexed Fiber Optic Sensors
(A. D. Kersey & J. P. Dakin, eds), Proceedings Volume 1586,
International Society for Optics and Photonics
,
Bellingham, WA, USA
, pp.
117
129
.
Lai
W. W. L.
Chang
R. K. W.
Sham
J. F. C.
Pang
K.
2016
Perturbation mapping of water leak in buried water pipes via laboratory validation experiments with high-frequency ground penetrating radar (GPR)
.
Tunnelling and Underground Space Technology
52
,
157
167
.
Li
J.
Zhao
Z.
2001
Hydraulics (Volume II)
.
Hohai University Press
,
Nanjin, China
.
Li
Q.
Du
X.
Zhang
H.
Li
M.
Ba
W.
2018
Liquid pipeline leakage detection based on moving windows LS-SVM algorithm
. In:
33rd Youth Academic Annual Conference of Chinese Association of Automation (YAC), 2018
,
IEEE
,
Piscataway, NJ, USA
, pp.
701
705
.
Liggett
J. A.
Chen
L. C.
1994
Inverse transient analysis in pipe networks
.
Journal of Hydraulic Engineering
120
(
8
),
934
955
.
Lim
J.
2015
Underground pipeline leak detection using acoustic emission and crest factor technique
. In:
Advances in Acoustic Emission Technology
(G. Shen, Z. Wu & J. Zhang, eds),
Springer
,
New York, USA
, pp.
445
450
.
Lu
W.
Liang
W.
Zhang
L.
Liu
W.
2016
A novel noise reduction method applied in negative pressure wave for pipeline leakage localization
.
Process Safety and Environmental Protection
104
,
142
149
.
Meniconi
S.
Brunone
B.
Ferrante
M.
Capponi
C.
Carrettini
C.
Chiesa
C.
Segalini
D.
Lanfranchi
E.
2015
Anomaly pre-localization in distribution–transmission mains by pump trip: preliminary field tests in the Milan pipe system
.
Journal of Hydroinformatics
17
(
3
),
377
389
.
Miller
R. K.
Pollock
A. A.
Watts
D. J.
Carlyle
J. M.
Tafuri
A. N.
Yezzi
J. J.
Jr
1999
A reference standard for the development of acoustic emission pipeline leak detection techniques
.
NDT & E International
32
(
1
),
1
8
.
Nakhkash
M.
Mahmood-Zadeh
M. R.
2004
Water leak detection using ground penetrating radar
. In:
Proceedings of the Tenth International Conference on Ground Penetrating Radar, 2004. GPR 2004
(E. C. Slob, A. G. Yarovoy & J. B. Rhebergen, eds),
IEEE
,
Piscataway, NJ, USA
, pp.
525
528
.
Ni
L.
Jiang
J.
Pan
Y.
2013
Leak location of pipelines based on transient model and PSO-SVM
.
Journal of Loss Prevention in the Process Industries
26
(
6
),
1085
1093
.
Qu
Z.
Feng
H.
Zeng
Z.
Zhuge
J.
Jin
S.
2010
A SVM-based pipeline leakage detection and pre-warning system
.
Measurement
43
(
4
),
513
519
.
Ridolfi
E.
Servili
F.
Magini
R.
Napolitano
F.
Russo
F.
Alfonso
L.
2014
Artificial neural networks and entropy-based methods to determine pressure distribution in water distribution systems
.
Procedia Engineering
89
,
648
655
.
Simpson
A.
Murphy
L.
Dandy
G.
1993
Pipe network optimisation using genetic algorithms
. In:
Proceedings, Specialty Conference of Water Resources Planning and Management Division, American Society of Civil Engineers, Seattle, Washington, USA, May
, pp.
392
395
.
Simpson
A. R.
Dandy
G. C.
Murphy
L. J.
1994
Genetic algorithms compared to other techniques for pipe optimization
.
Journal of Water Resources Planning and Management
120
(
4
),
423
443
.
Smith
L. A.
2010
Options for Leak and Break Detection and Repair for Drinking Water Systems
.
Battelle Press
,
Columbus, OH, USA
.
Stampolidis
A.
Soupios
P.
Vallianatos
F.
Tsokas
G. N.
2003
Detection of leaks in buried plastic water distribution pipes in urban places – a case study
. In:
Proceedings of the 2nd International Workshop on Advanced Ground Penetrating Radar, 2003
(A. Yarovoy, ed.),
IEEE
, pp.
120
124
.
Sun
L.
Chang
N.
2014
Integrated-signal-based leak location method for liquid pipelines
.
Journal of Loss Prevention in the Process Industries
32
,
311
318
.
Vítkovský
J. P.
Simpson
A. R.
Lambert
M. F.
2000
Leak detection and calibration using transients and genetic algorithms
.
Journal of Water Resources Planning and Management
126
(
4
),
262
265
.
Wang
X.
Chen
Z.
Zhong
X.
Lu
N.
2018
PSO-SVM based leakage diagnosis method of water supply pipeline
.
Modern Electronics Technique
41
(
7
),
156
159, 164
.
Watters
G. Z.
1984
Analysis and Control of Unsteady Flow in Pipelines
.
Butterworths
,
Boston, MA, USA
.
Yang
J.
Wen
Y.
Li
P.
2010
Approximate entropy-based leak detection using artificial neural network in water distribution pipelines
. In:
11th International Conference on Control Automation Robotics & Vision, 2010
,
IEEE
,
Piscataway, NJ, USA
, pp.
1029
1034
.
Zhang
S.
Fu
L.
2002
Operation and management of long-distance pipeline in Baoji
.
China Water & Wastewater
2
,
86
87
.
Zhang
R.-H.
Ni
M.
2004
Principle and applications of the fiber optic hydrophone
.
PHYSICS-BEIJING
33
(
7
),
503
507
.
Zhang
D.
Zhang
C.
2008
Reasons and countermeasures for break accident of wide-diameter prestressed concrete water transmission pipe in Baoji City
.
China Water & Wastewater
10
,
99
102
.
Zhang
X.
Gao
B.
Song
Z.
2003
The research of acoustic measuring of gas temperature employing cross-correlation algorithm
.
Proceedings of the Chinese Society for Electrical Engineering
23
(
4
),
185
188
.
Zhang
Y.
Chen
X.
Zhang
S.
Li
X.
Zhang
M.
Lu
H.
2006
Study on the causes and countermeasures of pipe explosion of large diameter water pipeline in Changzhou
.
Water & Wastewater Engineering
7
,
89
93
.