applied

sciences
Article
Application of FBG Based Sensor in Pipeline
Safety Monitoring
Tao Jiang 1 , Liang Ren 1, *, Ziguang Jia 2 , Dongsheng Li 1 and Hongnan Li 1,3
1 School of Civil and Hydraulic Engineering, Dalian University of Technology, Dalian 116024, China;
jiangtao@mail.dlut.edu.cn (T.J.); dsli@dlut.edu.cn (D.L.); hnli@dlut.edu.cn (H.L.)
2 School of Ocean Science and Technology, Dalian University of Technology, Panjin 124221, China;
jiaziguang@dlut.edu.cn
3 School of Civil Engineering, Shenyang Jianzhu University, Shenyang 110168, China
* Correspondence: renliang@dlut.edu.cn; Tel.: +86-411-8470-6384

Academic Editor: Chien-Hung Liu

Received: 29 March 2017; Accepted: 12 May 2017; Published: 24 May 2017

Abstract: Pipeline leakage and corrosion are two serious threats to pipeline safety operation.
Therefore, to ensure the safety operation of long-distance pipeline, it is of great significance to
conduct pipeline monitoring. Since hoop strain is an effective indicator to reflect the inner pressure
fluctuation and the wall thickness reduction of the pipeline, a method of monitoring leakage and
corrosion simultaneously was proposed based on hoop strain measurement. In order to test the hoop
strain variation, this paper introduces a fiber Bragg grating (FBG) strain hoop sensor. To verify the
monitoring method and the performance of this FBG strain sensor, a pipeline leakage simulation
experiment and corrosion simulation experiment were conducted on an actual pipeline and a
polyvinyl chloride (PVC) pipe model, respectively. The experimental results demonstrate that
the pipeline leakage and corrosion can be detected by the FBG hoop strain sensor. The FBG strain
hoop sensor is a promising device in pipeline safety monitoring.

Keywords: FBG; pipeline; leakage; corrosion; monitoring

1. Introduction
Pipeline is an important means to transport oil and gas through long distance [1,2]. However, due
to the effect of the inner fluids and the harsh environmental conditions, these pipelines are particularly
prone to be corroded [3]. Once the pipes are subjected to corrosion, it may probably cause leakage
or breakages [4]. Thus, a number of pipeline detection technologies have been developed to ensure
pipeline safety. Pole climbing robots that perform periodical inspections with non-destructive testing
probes are widely used for assessing the progression of material degradation and detecting welding
defects [5]. In order to inspect 80–100 mm pipelines in an indoor pipeline environment, Kwon et al.
design a motion planning algorithm of a caterpillar-based pipeline robot [6]. An ultrasonic inspection
robot equipped with an electromagnetic acoustic transducer was developed for inspection of a steel
pipeline in service [7]. However, inspection techniques cannot provide real-time monitoring of the
pipelines. Therefore, leakage or pipeline defects may not be detected in time and may cause much
larger economic loss and environmental pollution [8]. Consequently, it is more significant to perform
permanent and real-time integrity monitoring.
Generally, oil and gas pipelines are buried underground and extend hundreds of kilometers. These
working conditions mean that many traditional structural monitoring technologies are not applicable
for pipeline structure. A fiber optic sensor, with its superior immunity to electromagnetic interference,
long distance transmission, good embeddability and reliability [9], is an ideal approach for pipeline
safety monitoring. Optical fiber based innovative distributed temperature monitoring techniques have

Appl. Sci. 2017, 7, 540; doi:10.3390/app7060540 www.mdpi.com/journal/applsci

Appl. Sci. 2017, 7, 540 2 of 12

proven to be an efficient way to detect and localize leakages along pipelines [10]. Since leakages from the
pipeline lead to regional temperature anomalies in the vicinity of the pipeline, Brillouin-based distributed
temperature monitoring systems have been developed and equipped with a 55 km pipeline for the leakage
localization [11]. Zou et al. developed a Brillouin-scattering-based fiber optic distributed strain sensor and
applied it to measure the longitudinal and hoop strain in a steel pipe model with wall thinning defects.
According to the strain distribution, the locations of structural defects were found [12]. To detect the
internal corrosion of pipeline, a novel distributed sensor, employing metal coated polymer-clad fiber optic
cables, was studied and developed theoretically by Alhandawi et al. [13]. However, the widely applied
distributed optic fiber sensor has its own advantages and limitations. For example, optical time-domain
reflectometer (OTDR) based technique has distributed measurement over a long distance due to its larger
dynamic range, while its spatial resolution is usually on the order of meters to tens of meters [14], and its
measurement accuracy is not high. Generally, the distributed sensors are more suitable for localizing
damage or leakage instead of measuring the degree of damage or corrosion. An FBG sensor, with the
special advantages of high measurement accuracy [15], is also suitable for pipeline monitoring [16].
To monitor pipeline leakage, Jia et al. applied FBG sensors to measure hoop strain at regular points along
the pipeline [17]. Felli et al. presented FBG accelerometers for detecting fuel theft from pipelines [18].
Zhou et al. applied the FBG sensors in the monitoring of the dynamic response of a submarine pipeline
model under a variety of dynamic loading conditions [19]. Lee et al. proposed a method to monitor
the damage of pipeline using FBG sensors to measure guided wave responses [20]. In order to detect
and localize pipeline leakage, Hou et al. utilized an FBG strain sensor in the natural gas pipeline model
to detect a negative pressure wave (NPW) caused by leakage [21]. It can be concluded that the current
study based on FBG sensors mainly focus on the pipeline damage detection and leakage monitoring.
However, few studies about corrosion monitoring have been conducted, despite having the advantage
of high measurement accuracy. Moreover, the existing optic fiber based monitoring technique cannot
conduct corrosion monitoring and leakage monitoring simultaneously. In this paper, an FBG hoop strain
sensor is proposed to monitor pipeline corrosion and leakage simultaneously. Two model experiments
were conducted to test the performance of an FBG strain hoop sensor. To verify the correctness of the
monitoring theory, a numerical analysis was also carried out using SAP2000 software (V15, Computers
and Structures, Inc., New York, NY, USA).

2. Pipeline Safety Monitoring Theory

To implement pipeline leakage and corrosion simultaneous monitoring, a key parameter is
required to reflect their respective characteristics. Thus, a hoop strain theory is proposed.

2.1. Hoop Strain Theory

Inner pressure causes the pipeline expansion in the radial direction. Based on the theory of
Material Mechanics, the relationship between the hoop strain, the inner pressure and the wall thickness
of the pipeline, is given directly:
ε = P × R/E × t (1)

where ε is the hoop strain of the pipeline; P is the applied pressure; E is the Young’s modulus and t is
the wall thickness of the pipeline.

2.2. Leakage Monitoring

When pipeline leakage occurs, the fluid suddenly escapes from the leakage point, causing the
instantaneous pressure drop. Simultaneously, an NPW is generated and propagate toward both ends
of the pipeline at a certain velocity [17]. Traditionally, several pressure sensors are mounted along the
pipeline to detect the NPW. According to Equation (1), the hoop strain is proportional to the inner
pressure variation, since t and R can be considered as the constants during the pipeline leakage. Thus,
pipeline leakage can be detected via the hoop strain measurement.
 Appl. Sci. 2017, 7, 540 3 of 12

2.3.2017,
Appl. Sci. Corrosion
7, 540 Monitoring 3 of 12

For pipeline corrosion monitoring, it is assumed that P and R are the constants since pipelines
2.3. Corrosionwork
usually in the condition of steady pressure and the radius of the pipeline is a constant. Therefore,
Monitoring
the hoop strain is inversely proportional to the wall thickness of pipeline based on Equation (1).
For pipeline
According tocorrosion monitoring,
this theory, it is assumed
the wall thickness that Pwhich
reduction, and Risare
thethe constants
most since pipelines
direct phenomenon caused
usually work in the condition of steady pressure and the radius
by corrosion, can be reflected by the hoop strain directly. of the pipeline is a constant. Therefore,
the hoop strain is inversely proportional to the wall thickness of pipeline based on Equation (1).
According to thisand
2.4. Leakage theory, the wall
Corrosion thickness reduction, which is the most direct phenomenon caused by
Distinction
corrosion, can be reflected by the hoop strain directly.
Both leakage and corrosion will cause the hoop strain variation. Generally, pipeline leakage
occurs within
2.4. Leakage a few seconds,
and Corrosion and the instantaneous pressure drop causes the hoop strain to decrease.
Distinction
However, for pipeline corrosion, which directly leads to the pipeline wall thickness reduction, will induce
Both leakage and corrosion will cause the hoop strain variation. Generally, pipeline leakage occurs
the hoop strain increase in a few months or years. Thus, leakage and corrosion can be distinguished
within a few seconds, and the instantaneous pressure drop causes the hoop strain to decrease. However,
by the hoop strain decrease or increase.
for pipeline corrosion, which directly leads to the pipeline wall thickness reduction, will induce the
Based on the theory of hoop strain monitoring, pipeline leakage and the corrosion level can be
hoop strain increase in a few months or years. Thus, leakage and corrosion can be distinguished by
measured. Therefore, whether the hoop strain can be measured accurately is very crucial. An FBG
the hoop strain decrease or increase.
strain hoop sensor, which can be attached on the external surface of the pipeline and measure
Based on the theory of hoop strain monitoring, pipeline leakage and the corrosion level can be
the overall circumferential deformation on the same cross-section, is designed and applied in the
measured. Therefore, whether the hoop strain can be measured accurately is very crucial. An FBG
model tests.
strain hoop sensor, which can be attached on the external surface of the pipeline and measure the overall
circumferential deformation on the same cross-section, is designed and applied in the model tests.
3. Sensor Introduction
3. Sensor Introduction
3.1. Sensor Design
3.1. SensorFigures
Design 1 and 2 present the schematic diagram and the photo of FBG strain hoop sensor,
respectively.
Figures 1 and This sensor isthe
2 present comprised
schematic of diagram
an FBG sensor,
and thetwo capillary
photo of FBG copper tubes,
strain hoopthree gripper
sensor,
respectively. This sensor is comprised of an FBG sensor, two capillary copper tubes, three gripper two
blocks and a steel tube. The fiber in both sides of an FBG is packaged with the epoxy resin in the
blockscapillary copper
and a steel tube.tubes.
The The
fibersteel tubesides
in both is used to protect
of an FBG is the fragile with
packaged bare fiber from damage.
the epoxy Meanwhile,
resin in the two
the steel tube is mounted on the outer surface of the pipeline, providing a
capillary copper tubes. The steel tube is used to protect the fragile bare fiber from damage. Meanwhile, track for the capillary
copper
the steel tubetubes slide inon
is mounted it.the
Theouter
two surface
gripperofblocks fixed onproviding
the pipeline, the steel tube arefor
a track temporarily
the capillaryjammed
copperin an
tubesanchorage
slide in it. member when bonding
The two gripper the steel
blocks fixed tube
on the to the
steel tubeouter surface of the
are temporarily pipeline,
jammed ensuring
in an that the
anchorage
steel tube can be closely mounted onto the pipeline surface. After bonding
member when bonding the steel tube to the outer surface of the pipeline, ensuring that the steel tube the steel tube to the
can bepipeline, the two gripper
closely mounted onto theblocks
pipelineat the movable
surface. Afterend and the
bonding thefixed
steelend
tubeare jammed
to the in an
pipeline, anchorage
the two
member (as shown in Figure 2). The movable end shifts position when
gripper blocks at the movable end and the fixed end are jammed in an anchorage member (as showna circumferential deformation
happens,
in Figure 2). Thewhich, in turn,
movable enddrags
shiftsthe gripper
position block,
when causing deformation
a circumferential in the happens,
deformation grating area and the
which,
gripper
in turn, dragstubes.
the gripper block, causing deformation in the grating area and the gripper tubes.

bare fiber and FBG

steel tube
d1

d2

capillary copper tube

epoxy resin
gripper block

fixed end
movable end

df

Figure
Figure 1. Schematic
1. Schematic diagram
diagram of a fiber
of a fiber BraggBragg grating
grating (FBG)(FBG)
strainstrain
hoop hoop sensor.
sensor.
If the fiber packaged in the FBG strain hoop sensor is not tightened under working conditions,
If the fiber packaged in the FBG strain hoop sensor is not tightened under working conditions,
there may be no wavelength variation when pipeline leakage happens during its service. Therefore,
there may be no wavelength variation when pipeline leakage happens during its service. Therefore,
a pre-tension mechanism (presented in Figure 3) was developed to load pre-stressing to the fiber and
a pre-tension mechanism (presented in Figure 3) was developed to load pre-stressing to the fiber and
fix the FBG strain hoop sensor simultaneously. In practical application, the wavelength variation
Appl. Sci.
Appl. Sci. 2017,
2017, 7,
7, 540
540 44 of
of 12
12
Appl. Sci.
Appl. Sci. 2017,
2017, 7,
7, 540
540 44 of
of 12
12

fix the
fix the FBG strain
strain hoop sensor
sensor simultaneously. InIn practical application,
application, the wavelength
wavelength variation
fix the FBG
FBG strain hoop
hoop sensor simultaneously.
simultaneously. In practical
practical application, thethe wavelength variation
variation
caused by
caused by the
thepre-tension
pre-tensionsystem
systemisisis recommended
recommended to be restricted under 111 nm
nm so that
that the FBG will
caused by
caused by the
the pre-tension
pre-tension system
system isrecommended
recommended toto
bebe
to restricted
restricted
be under
under
restricted 1 nm
under so
so that
nm the the
so that FBGFBG
the willwill
FBG not
will
not
not
be be
be easily
easily
easily destroyed
destroyed
destroyed during
during
during testing.
testing.
testing.
not be easily destroyed during testing.

Figure 2. Photo
Figure of FBG
2. Photo
Photo strain
of FBG
FBG hoop
strain sensor.
hoop sensor.
Figure 2. of strain hoop sensor.
Figure 2. Photo of FBG strain hoop sensor.

Figure
Figure
Figure 3. Photo
3. Photo
3. Photo of pre-tension
of of pre-tension
pre-tension system.
system.
system.
Figure 3. Photo of pre-tension system.

3.2. Sensor
3.2.
3.2. Sensor Calibration
Sensor Calibration Test
Calibration Test
Test
3.2. Sensor Calibration Test
To
To characterize
characterize the
the workingperformance
working performance of the FBG strain hoop sensor, the calibration
calibration tests
To
To characterize
characterizethe working performance
theworking performance of of
thethe
of FBG
the FBG
FBG strain
strain hoop
strain hoop sensor,
sensor,
hoop the
the calibration
sensor, tests
tests were
the calibration tests
were
were conducted
conducted
conducted on a on
on
steel a
a
pipesteel
steel pipe
pipe
model model
model
with the with
with
outer the
the outer
outer
diameter diameter
diameter
219 mm 219
219
and mm
mm
wall and
and wall
wall
thickness 6 thickness
thickness
mm, which 6
6 mm,
mm,
is the
were conducted on a steel pipe model with the outer diameter 219 mm and wall thickness 6 mm,
which
which
same
is
is the
the same
same as
as the
the actual
actual pipeline
pipeline used
used for
for experiment.
experiment. The
The test
test model
model is
is presented
presented in
in Figure
Figure 4.
4.
whichasisthe theactual
same as pipeline usedpipeline
the actual for experiment.
used for The test model
experiment. Theis test
presented
model in Figure 4. The
is presented internal
in Figure 4.
The internal
The internal
pressure was
pressure
pressure
loaded by
was
wasa
loaded
loaded
water
by aa water
by
pump
waterapump
with
pump
maximum
withpressure
with a maximum
a maximum of
pressure
pressure
1.5 MPa.
of 1.5
of 1.5 MPa.
During
MPa.
the
During
During
test, the
the
the
pipe
The internal pressure was loaded by a water pump with a maximum pressure of 1.5 MPa. During the
test, the
test, the pipe
pipeloaded
modelstep
was by loaded step by step
step continuously from 000 KPa
KPa to 500 KPa atat an interval of 100
100 KPa.
model pipe model
test, thewas model was
was loaded step step
loaded by
by step continuously
continuously
step from 0 KPafrom
continuously to 500
from KPato at500
to KPa
KPa at an
an interval
500 ofinterval
an 100 KPa.
interval of
of 100 KPa.
KPa.

Figure 4.
Figure 4. A
A pipeline
pipeline model
model used
used for
for aa calibration
calibration test.
test.
Figure
Figure 4. pipeline
4. A A pipeline model
model used
used forfor a calibration
a calibration test.
test.
The sensitive
The sensitive coefficient, reflecting
reflecting the relationship
relationship between the the pressure and and hoop strain,
strain, is aa
The sensitive coefficient,
coefficient, reflecting the the relationship between
between the pressure
pressure and hoop hoop strain, is is a
key parameter
key parameter
The sensitive for coefficient,
the FBG
FBG strain
strain hoop sensor.
reflecting sensor.
the As shown
shown
relationship in Figure
Figure
between the5,5, the sensitivity
sensitivity
pressure and hoop coefficient of
strain, isof the
a key
key parameter for the FBG strain hoop sensor. As shown in Figure 5, the sensitivity coefficient of the
for the hoop As in the coefficient the
FBG strain
parameter
FBG hoop
for the sensors
FBG is
strain9.79
hoopKPa/με
sensor.andAs the hoop
shown strain
in measured
Figure 5, the by the
sensitivity FBG strain
coefficient hoop
of sensor
the FBG
FBG strain
strain hoop
hoop sensors
sensors isis 9.79
9.79 KPa/με
KPa/με and and the
the hoop
hoop strain
strain measured
measured by by the
the FBG
FBG strain
strain hoop
hoop sensor
sensor
is linear
linear
strain
is to the
hoop the inneris
sensors pressure
9.79 KPa/µεfrom thethe
andfirst
the to
hoopthe strain
fifth load
load step, demonstrating
measured demonstrating
by the FBG strain that hoop
the FBG
FBG strain
sensor is
is linear to
to the inner
inner pressure
pressure from
from the first
first to
to the
the fifth
fifth load step,
step, demonstrating that that the
the FBG strain
strain
hoop sensor
linear
hoop sensor is sensitive
to the inner
is sensitive
pressureto from
to the pressure
the pressure variation.
the first to Therefore,
the fifth Therefore,
variation. leakage monitoring
load step, demonstrating
leakage monitoring is FBG
that the
is viable based
strain
viable based on
hoop
on
hoop sensor is sensitive to the pressure variation. Therefore, leakage monitoring is viable based on
an FBG
FBGisstrain
sensor
an sensitive hoopto sensor. Fourteen
the pressure FBG strain
variation. strain hoop leakage
Therefore, sensors’monitoring
sensitive coefficients
coefficients wereon
is viable based extracted
an FBG
an FBG strain
strain hoophoop sensor.
sensor. Fourteen
Fourteen FBG FBG strain hoop hoop sensors’
sensors’ sensitive
sensitive coefficients were were extracted
extracted
and shown
strain
and shown
hoop in Figure
sensor. 6. Obviously,
Obviously,
Fourteen the sensitive
FBG strain sensitive coefficients
hoop sensors’ fluctuate
sensitive around
coefficients weretheextracted
value of of and
10 KPa/με,
KPa/με,
shown
and shown in in Figure
Figure 6.6. Obviously, the the sensitive coefficients
coefficients fluctuate
fluctuate around
around thethe value
value of 10 10 KPa/με,
indicating
in Figure
indicating that
6. the
Obviously,FBG strain
the hoop
sensitive sensor has
coefficients stable performance
fluctuate around since
the the
value packaging
of 10 KPa/µε, mechanism
indicating of
indicating that
that thethe FBG
FBG strain
strain hoop
hoop sensor
sensor hashas stable
stable performance
performance since since the
the packaging
packaging mechanism
mechanism of of
the
that
the sensor
the FBG has no
strain adverse
hoop effect
sensor on
has the hoop
stable strain
performance measurement.
since the packaging mechanism of the sensor
the sensor
sensor has
has no no adverse
adverse effect
effect on
on the
the hoop
hoop strain
strain measurement.
measurement.
has no adverse effect on the hoop strain measurement.
Appl.
Appl. Sci. 2017,
2017, 7, 540
540 55 of 12
Appl. Sci.
Sci. 2017, 7,
7, 540 5 of
of 12
12

Equation
Equation yy =
= aa +
+ b*
b*
500
500 Adj.
Adj. R-Squar
R-Squar 0.99055
0.99055
Slope
Slope 9.79289
9.79289

400
400

Pressure(KPa)
Pressure(KPa)
300
300

200
200

100
100

00

00 10
10 20
20 30
30 40
40 50
50
Hoop
Hoop strain
strain ((με
με))

Figure
Figure
Figure 5.
5. Relationship
5. Relationship
Relationship between
between thethe
between pressure
pressure
the andand
pressure hoop
hoop
and strain.
strain.
hoop strain.

15
15

12
με))

12
(KPa/με
coefficient(KPa/

99
Sensitivecoefficient

66
Sensitive

33

00
00 22 44 66 88 10
10 12
12 14
14 16
16
Sensor
Sensor number
number

Figure
Figure
Figure 6.
6. Calibration
testtest
Calibration
6. Calibration test results.
results.
results.

4.
4. Experimental Pipeline
Experimental Pipeline and
Pipeline and Sensor
and Sensor Installation
Sensor Installation
Installation
4. Experimental

4.1. Experimental
4.1. Pipeline Introduction
Experimental Pipeline Introduction
Introduction
The pipeline
The pipeline leakage
leakage experiment
experiment was was conducted
conducted on on part
part of
of the
the fire-fighting pipeline
fire-fighting pipeline
fire-fighting pipeline withwith
with aa total
total
length
length of
length of 600 m (shown
of 600 m (shown in Figures
(shown in Figures 7 and
Figures 77 and 8).
and 8). The outer diameter and the wall thickness of
8). The outer diameter and the wall thickness of the pipeline the pipeline
are
are 219
219 mm
mm and
and 6
6 mm,
mm, respectively.
respectively.
are 219 mm and 6 mm, respectively. The pipeline The
The pipeline
pipeline leakage
leakage
leakage was
was simulated
simulated
was simulated by
by opening
by opening opening the
the existing
the existing existing
valves
valves S1–S11,
valves S1–S11,
S1–S11, whose
whose whose
positionspositions
positions are displayed
are displayed
are displayed in Figurein Figure
in Figure 7. The actual
7. The valve
7. The actual valve
actualisvalve is presented
is presented
presented in Figurein Figure
in Figure 8b.
8b.
8b. A fire
A
A fire
fire service
service pump pump
service pump
was usedwas
wastoused loadto
used load
load pressure
topressure pressure
into the into
into the
the pipeline.
pipeline. pipeline. Fourteen
Fourteen
Fourteen FBG hoopFBG hoop
hoop strain
FBG strain strain
sensors sensors
sensors
were
were
were mounted
mounted on the on
mounted on the
the pipeline
pipeline pipeline
model with model
modelthe with
with the
same the same
same
interval of interval
interval of
of 20
20 m, forming 20 am,
m, forming
arrayaa (illustrated
forming
sensor sensor
sensor array
array
in
(illustrated
(illustrated
Figure in
7). It in Figure
Figure
should 7).
be 7).
notedIt should
It should be noted
that thebepipeline that
noted that the
between pipeline
the pipeline
sensors L1between
between sensors
and L14sensors L1
L1 and L14
and
is the experiment is
is the
L14section
the
experiment
experiment
with the lengthsection
section with
with
of 240 m.theThelength
the effectof
length of240
of 240 m.
m. The
The effect
temperature effect of
of temperature
on the temperature
FBG strain hoopon
on the
the FBG
FBG strain
sensor strain hoop
hoop sensor
is negligible sensor
since
is negligible
is negligible
pipeline since
since
leakage pipeline
pipeline
occurs within leakage
leakage occurs
a fewoccurs
seconds.within
within a few
a few
The four FBGseconds.
seconds. The four
The four
temperature FBG temperature
FBG temperature
sensors, mounted close sensors,
sensors,
to the
mounted
mounted
sensors L1,close
close to
to the
L4, L8 andsensors
the sensors
L14, were L1, L4,
L4, L8
L8 and
L1, utilized and L14,
L14, were
to achieve were utilized
utilized to
temperature achieve
achieve temperature
tocompensationtemperature compensation
compensation
of the four FBG strain
of
of the
hoop four
four FBG
thesensors FBGin strain
strain hoop sensors
sensors in
hoopmonitoring.
long-term in long-term
long-term monitoring.
monitoring.
Appl. Sci. 2017, 7, 540 6 of 12
Appl. Sci. 2017, 7, 540 6 of 12
Appl. Sci. 2017, 7, 540 6 of 12

Appl. Sci. 2017, 7, 540 6 of 12

Figure 7. Schematic
Schematic diagram of pipeline model and sensor position.
Figure 7.7.
Figure Schematicdiagram
diagramof
of pipeline
pipeline model and sensor
model and sensorposition.
position.

Figure 7. Schematic diagram of pipeline model and sensor position.

Valve
Valve

Pipeline
Pipeline Valve

Pipeline

(a)
(a) ((bb))
(a)
Figure 8. Picture
Figure 8. Picture
(b) (b) valve.
of actual pipeline; (a) pipeline;
Picture of actual pipeline; (a) pipeline; (b) valve.
Figure 8. Picture of actual pipeline; (a) pipeline; (b) valve.
4.2.4.2. Sensor Installation
Sensor Installation
4.2. Sensor
BeforeInstallation
installing the FBG strain hoop sensor, a sander and abrasive paper were used to burnish
Before installing
installingthe theFBGFBGstrain
strainhoop
hoopsensor,
sensor,a sander
a sander andand abrasive
abrasive paper
paper were
were used used to burnish
to burnish the
the area for sensor
Before installinginstallation,
the FBG providing
strain hoop asensor,
smoothaand uniform
sander and surface. After
abrasive paper burnishing,
were used the
to surface
burnish
the
area area for
for cleanedsensor installation,
sensor installation, providing
providing in a
a smoothsmooth and and uniform
uniform surface.
surface. After burnishing,
After burnishing, the
theanchoragesurface
surface was
was
the area for by cotton
sensor immersed
installation, providing alcohol.
a smooth Afterand cleaning
uniform the pipeline
surface. After surface,
burnishing, the theanchorage
surface
was cleaned
cleaned by by
cotton cotton
immersed immersed
in in alcohol.
alcohol. After After
cleaning cleaning
the pipelinethe pipeline
surface, the surface,
anchorage themember was
member
was was by
cleaned bonded
cotton onimmersed
the pipeline in surface.
alcohol. Then,
After the gripperthe
cleaning block connected
pipeline withthe
surface, theanchorage
steel tube
member
bonded was
on
was stretchedthebonded
pipeline
along the
on the pipeline
surface. Then,
circumferential
surface.
the Then,
gripper
direction
the
block
in the
order
gripper
connected
to make
block
with connected
the
theconnected steel
steel tube with
with
tube
seamlessly
the
was steel tube
stretched
contact
wasmember
stretched wasalong
bonded theon the pipeline surface.
circumferential direction Then,inpipeline
order gripper block the steel tube
along
the the
pipe circumferential
surface. After direction
bonding theinsteel
order to to
tube makethe the steeltosurface,
make
tube the steel tube
seamlessly
the contact
pre-tension seamlessly
the pipe
system contact
surface.
switched
was stretched along the circumferential direction in order to make the steel tube seamlessly contact
theto
Afterpipe surface.
bonding thetheAfter
steelbonding
tube the
thesteel
blocktoconnected tube
pipeline tothe
the pipeline
withsurface, thefiber, surface,
pre-tension the pre-tension
system switched system
the to switched
stretch
fiber.the
thestretch
pipe surface. gripper
After bonding the steel tube to theoptic
pipeline applying
surface, pre-stress
the pre-tension tosystem optic
switched
to stretch
gripper
Finally, the gripper
block connected block connected
with theboth with
opticmovable the
fiber, applying optic fiber,
pre-stress applying
to the pre-stress
optic to
fiber. Finally,the optic fiber.
the gripper
to stretchthe thegripper
gripperblocks block at connected with the endoptic
and fixed applying
fiber, end were clamped
pre-stress toin the
the optic
anchorage
fiber.
Finally,
blocks atthe
member. both gripper
The FBGblocks
movable end at
strain and
hoopboth
fixed movable
endand
sensor were end
the and fixed
clamped
anchorage in theend were clamped
anchorage
member member. in The
theFigure
anchorage
FBG strain
Finally, the gripper blocks at both movable end and fixed end wereare presented
clamped in the in anchorage 9a.
member.
hoopTo sensor
preventTheand FBG
the the
FBG strain
anchorage
strain hoopmember
sensorfrom
sensor and
are the anchorage
presented
corrosion, a in member
Figure
specially 9a. To
designed are presented
prevent
steel the
box is FBG in strain
used Figure
to 9a.
hoop
protect
member. The FBG strain hoop sensor and the anchorage member are presented in Figure 9a.
To prevent
theprevent
sensor
To
the
pre-tension FBG
from corrosion,
the FBG
strain
system hoop
and
a specially
strain hoop
sensor
the sensor
FBG from
strain
designed corrosion,
hoop
fromsteel sensor,
box
corrosion,
a specially
as shown
is aused designed
in Figure
to protect
specially designed
steel box is
9b. box is used
the steel
pre-tension used
system to protect
and the
to protect
the
FBG pre-tension
strain hoop system
sensor, and
as the
shown FBG in strain
Figure hoop
9b. sensor,
the pre-tension system and the FBG strain hoop sensor, as shown in Figure 9b. as shown in Figure 9b.

(a) (b)
(a)
Figure 9. FBG strain sensor(a)
monuted (b)
(b) hoop sensor and anchorage
on pipeline surface: (a) FBG strain
member; (b) steel box.
Figure 9. FBG strain sensormonuted
Figure 9. FBG strain sensor monutedon onpipeline
pipeline surface:
surface: (a) FBG strain hoop sensor and anchorage
(a) FBG strain hoop sensor and anchorage
member; (b) steel box.
member; (b) steel box.
Appl. Sci.
Appl. Sci. 2017, 7, 540
2017, 7, 540 77 of
of 12
12

5. Experimental Results
5. Experimental Results
5.1. Leakage Monitoring
5.1. Leakage Monitoring
Due to limitations of the water pump, the actual pipeline model cannot simulate the leakage
Due to limitations of the water pump, the actual pipeline model cannot simulate the leakage
under the flow state. Prior to starting the leakage experiment, the pipeline was filled with water and
under the flow state. Prior to starting the leakage experiment, the pipeline was filled with water and
pressurized to 1.2 MPa. During the leakage experiment, the inner pressure dropped from 1.2 MPa to
pressurized to 1.2 MPa. During the leakage experiment, the inner pressure dropped from 1.2 MPa to
0 MPa by opening the valve suddenly. The Micron Optics SM130 (Micron Optics, Inc., Atlanta, GA,
0 MPa by opening the valve suddenly. The Micron Optics SM130 (Micron Optics, Inc., Atlanta, GA,
USA) was used to demodulate the wavelength of an FBG strain hoop sensor. The sampling rate in
USA) was used to demodulate the wavelength of an FBG strain hoop sensor. The sampling rate in this
this leakage experiment is 1000 Hz. Figure 10 presents the wavelength variation of sensor L1 when
leakage experiment is 1000 Hz. Figure 10 presents the wavelength variation of sensor L1 when the
the leakage was simulated by opening valve S1. It can be seen that the wavelength remains constant
leakage was simulated by opening valve S1. It can be seen that the wavelength remains constant before
before the leakage occurrence. However, when the leakage happened, the wavelength detected by
the leakage occurrence. However, when the leakage happened, the wavelength detected by the FBG
the FBG strain hoop sensor suddenly declined. This is due to the NPW propagating along the pipeline
strain hoop sensor suddenly declined. This is due to the NPW propagating along the pipeline model,
model, causing the inner pressure to suddenly drop, simultaneously leading to the hoop strain
causing the inner pressure to suddenly drop, simultaneously leading to the hoop strain decrease.
decrease. The time when the turning point (displayed in Figure 10) occurs can be considered as the
The time when the turning point (displayed in Figure 10) occurs can be considered as the arrival time
arrival time of NPW. The experiment results demonstrate that the wavelength detected by the FBG
of NPW. The experiment results demonstrate that the wavelength detected by the FBG strain hoop
strain hoop sensor can be taken as an effective indicator of leakage occurrence and the FBG strain
sensor can be taken as an effective indicator of leakage occurrence and the FBG strain hoop sensor is
hoop sensor is suitable for the pipeline leakage monitoring.
suitable for the pipeline leakage monitoring.

1529.20
L1

1529.15 Turning point

1529.10
Wavelength(nm)

1529.05

1529.00

1528.95

-2.5 0.0 2.5 5.0 7.5 10.0 12.5 15.0 17.5 20.0 22.5 25.0 27.5
Time(s)

Figure 10.
Figure Wavelength responses
10. Wavelength responses induced
induced by
by leakage
leakage point
point S1.
S1.

When leakage occurred

When leakage occurredatatS1, S1,the
thehoop
hoop strain
strain variation
variation of six
of six sensors
sensors (L2,(L2, L4, L8,
L4, L6, L6, L10
L8, and
L10 L13)
and
L13) with
with the same
the same intervalinterval
of 40 mofis40displayed
m is displayed
in Figurein Figure 11. Obviously,
11. Obviously, the lastthe last to
sensor sensor
detecttothe detect the
turning
turning point was L13, which is located the farthest from the leakage point
point was L13, which is located the farthest from the leakage point among the six sensors. However, among the six sensors.
However, for sensor
for sensor L2, which is L2, whichclosest
located is located closest
to the leakageto the leakage
point pointthe
S1 among S1 among the six
six sensors, sensors,
is the is the
first sensor
first sensor
to detect thetoturning
detect point.
the turning point. Inthe
In conclusion, conclusion,
closer thethe closerbetween
distance the distance betweenpoint
the leakage the leakage
and the
point
sensor,and
the the sensor,
earlier the earlier
the turning point theoccurs.
turning Thepoint occurs. results
experiment The experiment
demonstrate results
that demonstrate
NPW propagates that
NPW
along propagates
the pipelinealong
modelthe at pipeline
a certain model
speed. atThea certain
time whenspeed. The time
a turning when
point a turning
occurs can bepoint occurs
considered
can be considered as the arrival time of NPW. Based on the difference in time
as the arrival time of NPW. Based on the difference in time it takes for the NPW to reach the FBG it takes for the NPW to
reach the FBG strain hoop sensors, the leakage location can be calculated. Additionally,
strain hoop sensors, the leakage location can be calculated. Additionally, it can be seen from Figure 11 it can be seen
from Figure
that hoop 11 that
strain hoop of
variation strain variation
different of different
sensors under the sensors under the
same leakage same leakage
is different, is different,
the amplitude of
the amplitude of hoop strain variation decreases with the increase of the distance
hoop strain variation decreases with the increase of the distance between the leakage point and the between the leakage
point and the
FBG strain FBG
hoop strainThis
sensor. hoop sensor. Thisillustrates
phenomenon phenomenon illustrates
that the that the
NPW energy NPW energy
attenuates gradually attenuates
when
gradually when NPW
NPW propagates alongpropagates
the pipeline, along
and the pipeline,
perhaps thereand is a perhaps there is a certain
certain relationship between relationship
the NPW
between the NPW energy
energy attenuation and its attenuation
propagation anddistance.
its propagation distance. The
The experimental experimental
results suggest results suggest
that a leakage
that a leakage localization method according to the NPW energy attenuation is worthy of being
Appl. Sci. 2017, 7, 540 8 of 12

Appl. Sci. 2017, 7, 540 8 of 12

localization method according to the NPW energy attenuation is worthy of being studied. It can also
studied.
be It canfrom
concluded also Figure
be concluded
11 that from Figure of
the method 11 leakage
that the localization
method of leakage
based onlocalization
the NPW is based on the
limited by
NPW is limited
distance. by the
Therefore, distance. Therefore,
principle of NPWthe principle
energy of NPW
attenuation willenergy attenuation
be studied will be studied
comprehensively in the
comprehensively
next stage. in the next stage.

20 L13
L10
0 L8
L6
-20 L4
L2
-40
Hoop strain(με)

-60

-80

-100

-120

-140
-2.5 0.0 2.5 5.0 7.5 10.0 12.5 15.0 17.5 20.0 22.5 25.0 27.5
Time(s)
Figure 11. Hoop strain responses induced by leakage point S1.
Figure 11. Hoop strain responses induced by leakage point S1.

5.2. Corrosion Detection Results

5.2. Corrosion Detection Results
Prior to a pipeline corrosion experiment, a numerical simulation using SAP2000 software was
Prior to a pipeline corrosion experiment, a numerical simulation using SAP2000 software was
conducted to verify the pipeline corrosion monitoring theory based on hoop strain measurement.
conducted to verify the pipeline corrosion monitoring theory based on hoop strain measurement.
There are some basic assumptions in this numerical model: (1) the pipeline model was modeled by
There are some basic assumptions in this numerical model: (1) the pipeline model was modeled by a
a shell element; (2) the material is isotropic; (3) each part with different thickness is connected to
shell element; (2) the material is isotropic; (3) each part with different thickness is connected to another
another by a rigid connection; (4) the end restraint is selected as a hinge; and (5) the elasticity modulus
by a rigid 9connection; (4) the end restraint is selected as a hinge; and (5) the elasticity modulus is
is 210 × 10 Pa.
210 × 109 Pa.
The deformation nephogram generated by the SAP2000 software is shown in Figure 12 and the
The deformation nephogram generated by the SAP2000 software is shown in Figure 12 and the
wall thicknesses are marked by dimensions. To be more explicit, the scaling of nephogram is taken
wall thicknesses are marked by dimensions. To be more explicit, the scaling of nephogram is taken
as 2000, and the place circled in red is magnified five times. The numerical results are extracted and
as 2000, and the place circled in red is magnified five times. The numerical results are extracted and
displayed in Figure 13. Additionally, the linear equation between the hoop strain and the wall
displayed in Figure 13. Additionally, the linear equation between the hoop strain and the wall thickness
thickness of the numerical model is established. The numerical analytical results show a linear
of the numerical model is established. The numerical analytical results show a linear variation of the
variation of the hoop strain with a regression coefficient over 0.999, which demonstrates that the hoop
hoop strain with a regression coefficient over 0.999, which demonstrates that the hoop strain is in
strain is in inverse proportion to the wall thickness of pipeline strictly. It also can be seen that the
inverse proportion to the wall thickness of pipeline strictly. It also can be seen that the pipeline wall
pipeline wall extends uniformly and the circumferential deformations of the pipeline decrease
extends uniformly and the circumferential deformations of the pipeline decrease according to the rise
according to the rise of the wall thicknesses. The numerical results proved that the corrosion
of the wall thicknesses. The numerical results proved that the corrosion monitoring theory based on
monitoring theory based on hoop strain measurement is valid.
hoop strain measurement is valid.
Appl. Sci.Sci.
Appl. 2017, 7, 7,
2017, 540540 9 of 12 12
9 of
Appl. Sci. 2017, 7, 540 9 of 12

Figure 12. The deformation Nephogram generated by SAP2000 software.

Figure 12.
Figure 12. The deformation Nephogram
Nephogram generated
generated by
by SAP2000
SAP2000 software.
software.

44
44 Numerical results
Numerical results
40
40

36
36
Hoop strain(με)
Hoop strain(με)

32
32

28 y =130 x +0.21 R² = 0.999

28 y =130 x +0.21 R² = 0.999

24
24

0.18 0.20 0.22 0.24 0.26 0.28 0.30 0.32 0.34

0.18 0.20 0.22 0.24 0.26 0.28 -1 0.30 0.32 0.34
1/t(mm
-1 )
1/t(mm )
Figure Hoop
13.13.
Figure strain
Hoop extracted
strain from
extracted numerical
from results.
numerical results.
Figure 13. Hoop strain extracted from numerical results.

It It is difficult to achieve corrosion dates from the actual pipeline model within a short time since
It is
is difficult
difficult toto achieve
achieve corrosion
corrosion datesdates from
from the
the actual
actual pipeline
pipeline model
model within
within aa short
short time
time since
since
corrosion
corrosion is is
a a long-term
long-term process.
process. Therefore,
Therefore, inin order
order toto investigate
investigate thethe performance
performance of ofa a FBG
FBG strain
strain
corrosion is a long-term process. Therefore, in order to investigate the performance of a FBG strain
hoop sensorinin
hoop the pipeline corrosion detection,anan experiment was conducted ona PVC a PVC pipeline
hoop sensor
sensor in the the pipeline
pipeline corrosion
corrosion detection,
detection, an experiment
experiment was was conducted
conducted on on a PVC pipeline
pipeline
model
model since the PVC material is easy to be machined. This model (shown in Figure 14) consists of
model since the PVC material is easy to be machined. This model (shown in Figure 14) consists of
since the PVC material is easy to be machined. This model (shown in Figure 14) consists of
seven segments
seven with the same outer diameter of 250 mm but different wall thicknesses (e.g., 7.8mm,
mm,
seven segments
segments with with the
the same
same outer
outer diameter
diameter of of 250
250 mm
mm but but different
different wall
wall thicknesses
thicknesses (e.g.,
(e.g., 7.8
7.8 mm,
6.0
6.0 mm, 5.1 mm, 4.4 mm, 4.2 mm, 3.8 mm and 3.4 mm). The different wall thicknesses represent
6.0 mm,
mm, 5.1 5.1 mm,
mm, 4.44.4 mm,
mm, 4.24.2 mm,
mm, 3.8 3.8 mm
mm andand 3.4
3.4 mm).
mm). The The different
different wall
wall thicknesses
thicknesses represent
represent
differentuniform
different uniformcorrosion
corrosionlevels.
levels. FBG
FBG strain hoop sensors were mounted ononthethe
segments with wall
different uniform corrosion levels. FBG strain hoop sensors were mounted on the segments withwith
strain hoop sensors were mounted segments wall
thicknesses
wall thicknesses 5.1 mm,
5.1 4.44.4 mm,
mm, 4.44.24.2 mm,
mm, 4.2 3.8 mm and 3.4andmm, respectively. These sensors were connected
thicknesses 5.1 mm, mm, mm, 3.8mm,
mm3.8 andmm
3.4 mm, 3.4 mm, respectively.
respectively. These
These sensors sensors
were were
connected
to an interrogator
connected (Micron Optics
to an interrogator (Micron SM130)
Opticsthat sent data via thedata
Ethernet to Ethernet
a computer. aAn air pump
to an interrogator (Micron Optics SM130) thatSM130)
sent data that
viasent via to
the Ethernet thea computer. toAn computer.
air pump
Anwas used to provide a pressurized condition to simulate the working environment of a practical
wasairused
pump to was usedatopressurized
provide provide a pressurized
condition tocondition
simulatetothe simulate
working theenvironment
working environment
of a practicalof a
pipeline.
practical The demodulation instrument collected data from the FBG strain hoop sensors on the PVC
pipeline.pipeline. The demodulation
The demodulation instrument instrument
collectedcollected
data fromdata the from the FBG
FBG strain hoopstrain hoopon
sensors sensors
the PVCon
pipeline
the PVC model
pipeline with
model the internal
with the pressure
internal being
pressure loaded
being from
loaded 0 KPa
from to
0 50
KPa KPa.
to 50 KPa.
pipeline model with the internal pressure being loaded from 0 KPa to 50 KPa.
 Appl. Sci. 2017, 7, 540 10 of 12
Appl.Appl.
Sci. 2017, 7, 540
Sci. 2017, 7, 540 10 of10
12of 12

FBGFBG
strainstrain
hoophoop
sensor
sensor

Air pump
Air pump

Pressure sensor
Pressure sensor

Figure 14. Photo

Figure of the PVC
14. Photo pipeline model.
Figure 14. Photo of
of the
the PVC
PVC pipeline model.
pipeline model.

TheThehoophoop strain variation

strain measured
variation measured by five FBGFBG
by five strain hoop
strain hoopsensors is shown
sensors is shown in Figure
in Figure15. 15.
Because The
PVC hoop
is not strain
an variation
absolute linearmeasured
elastic by five leading
material, FBG strain to hoop
the hoop sensors
strain is shown under
variation in Figure
the 15.
Because PVC is not an absolute linear elastic material, leading to the hoop strain variation under the
Because
internal PVC isisnot
pressure anextremely
absolute linear elastic material, theleading to the hoop strain of variation under
internal pressure not is not extremely linearized.
linearized. Therefore,
Therefore, the regression
regressioncoefficient
coefficient the linear
of the linear
the internal
variations of the pressure
FBG is
strain not
hoopextremely
sensor islinearized.
0.975, less Therefore,
than 0.999. the regression
Nevertheless, coefficient
hoop strain of the
variationlinear
variations of the FBG strain hoop sensor is 0.975, less than 0.999. Nevertheless, hoop strain variation
andvariations
wallwall
and
of theofFBG
thickness
thickness the strain
pipeline
of the
hoopstillsensor
pipeline present
still
is as
present
0.975,
beinglessanthan
as being
0.999.
inverse
an inverse
Nevertheless,
proportion
proportion
hoop strain
relationship.
relationship.
variation
TheThehoop hoop
andmeasured
strains wall thicknessby the of the strain
pipeline stillsensors
presentgive as being an inverse proportion relationship. The hoop
strains measured by FBG
the FBG strain hoop hoop sensors the sensitivity
give the sensitivity1767 με/mm
1767 με/mmfor the
for PVC
the PVCpipeline
pipeline
strains
model, measured
proving by the
thatthat
the FBGFBG strain
strain hoop hoop sensors
sensors aregive the sensitivity
sensitive to the hoop1767 µε/mm
strain for thecaused
variations PVC pipeline
by by
model, proving the FBG strain hoop sensors are sensitive to the hoop strain variations caused
model,
corrosion. proving
Therefore, that the FBG
considering strain hoop
the the
good sensors
performanceare sensitive
of the to the
FBGFBG hoop
strain strain
hoop variations
sensor caused
in the
corrosion. Therefore, considering good performance of the strain hoop sensor in the
by corrosion.
experimental Therefore,
results, the hoopconsidering
strain the goodby
measured performance
the FBG of thehoop
strain FBG sensor
strain hoopcan sensorthe
reflect in the
experimental results, the hoop strain measured by the FBG strain hoop sensor can reflect the
experimental
corrosion degree results,
accurately.the hoop strain measured by the FBG strain hoop sensor can reflect the corrosion
corrosion degree accurately.
degree accurately.

540 540

510 510

480 480
Hoop strain(με)
Hoop strain(με)

450 450

420 420

390 390

360 360
y = 1767.7 x + 139.9
y = 1767.7 R2 =R0.975
x + 139.9 2 = 0.975
330 330
0.180.18 0.200.200.220.22 0.240.24 0.260.260.280.28 0.300.30
-1
1/t(mm ) -1)
1/t(mm

Figure 15. Results of the corrosion simulation experiment.

Figure 15. Results
Figure of the
15. Results of corrosion simulation
the corrosion experiment.
simulation experiment.

6. Conclusions
6. Conclusions
6. Conclusions
A method, based on hoop strain measurement, was proposed to detect pipeline corrosion and
A method,
A method, based on hoop
based on hoopstrain measurement,
strain measurement, waswasproposed to detect
proposed pipeline
to detect corrosion
pipeline andand
corrosion
leakage simultaneously. With the advantages of high sensitivity, good electrical-magnetic immunity
leakage simultaneously.
leakage simultaneously. With the the
With advantages
advantages of high sensitivity,
of high good
sensitivity, electrical-magnetic
good electrical-magnetic immunity
immunity
and chemical stability, an FBG strain hoop sensor was developed for the pipeline monitoring.
andand
chemical stability, an FBG strain hoop sensor was developed for the pipeline
chemical stability, an FBG strain hoop sensor was developed for the pipeline monitoring. monitoring. TwoTwo
Two model experiments including corrosion and leakage were conducted to verify the method and
model experiments including corrosion and leakage were conducted to verify
model experiments including corrosion and leakage were conducted to verify the method and the method and testtest
test the performance of this sensor. The leakage simulation experiment proved that a turning point
the the
performance
performance of this sensor.
of this TheThe
sensor. leakage simulation
leakage experiment
simulation proved
experiment thatthat
proved a turning point
a turning arose
point arose
duearose
to thedue
NPWto the NPW propagate
propagate along the along the pipeline.
pipeline. Meanwhile,Meanwhile,
pipeline pipeline
leakage leakage
can
due to the NPW propagate along the pipeline. Meanwhile, pipeline leakage can be detected by the be can be detected
detected by the by
FBGFBGstrain hoop
strain sensor.
hoop According
sensor. According to the corrosion
to the experiment,
corrosion experiment,FBGFBG strain hoop
strain hoopsensors were
sensors were
demonstrated
demonstrated to be
to sensitive to the
be sensitive hoop
to the strain
hoop variations
strain caused
variations by corrosion.
caused by corrosion.It can be concluded
It can be concluded
Appl. Sci. 2017, 7, 540 11 of 12

the FBG strain hoop sensor. According to the corrosion experiment, FBG strain hoop sensors were
demonstrated to be sensitive to the hoop strain variations caused by corrosion. It can be concluded
from the two experiments that the pipeline leakage and corrosion monitoring method based on hoop
strain measurement are practicable. Further investigation will be conducted to study the leakage
monitoring of the gas pipeline and improve the sensitivity of the FBG strain hoop sensor.

Acknowledgments: This work has been supported by the National Key Research and Development Program of
China (Grant No. 2016YFC0701107), the Fund of the National Natural Science Foundation of China (Grant Nos.
51421064, 51327003, 51678109 and 51608094). These grants are greatly appreciated.
Author Contributions: Tao Jiang performed the experiments and wrote the paper; Liang Ren and Tao Jiang
conceived and designed the experiments; Ziguang Jia analyzed the data; Dongsheng Li and Hongnan Li made
constructive comments on the experiments and paper writing.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Laurentys, C.; Bomfim, C.; Menezes, B.; Caminhas, W. Design of a pipeline leakage detection using expert
system: A novel approach. Appl. Soft Comput. 2011, 11, 1057–1066. [CrossRef]
2. Frings, J.; Walk, T. Distributed fiber optic sensing enhances pipeline safety and security. Oil Gas Eur. Mag.
2011, 3, 132–136.
3. Bartolini, C.; Cioccolanti, L.; Comodi, G.; Caresana, F. Distributed stirling engines for pipeline corrosion
protection. In Proceedings of the 15th International Stirling Engine Conference, Dubrovnik, Croatia,
27–29 September 2012.
4. Skjelvareid, M.; Birkelund, Y.; Larsen, Y. Internal pipeline inspection using virtual source synthetic aperture
ultrasound imaging. NDT E Int. 2013, 54, 151–158. [CrossRef]
5. Tavakoli, M.; Marques, L.; de Almeida, A. Development of an industrial pipeline inspection robot. Ind. Robot
Int. J. 2010, 37, 309–322. [CrossRef]
6. Kwon, Y.; Yi, B. Design and motion planning of a two-module collaborative indoor pipeline inspection robot.
IEEE Trans. Robot. 2012, 28, 681–696. [CrossRef]
7. Murayama, R.; Makiyama, S.; Kodama, M.; Taniguchi, Y. Development of an ultrasonic inspection robot
using an electromagnetic acoustic transducer for a Lamb wave and an SH-plate wave. Ultrasonics 2004, 42,
825–829. [CrossRef] [PubMed]
8. Sun, Z.; Wang, P.; Vuran, M.; Al-Rodhaan, M.; Al-Dhelaan, A.; Akyildiz, I. MISE-PIPE: Magnetic
induction-based wireless sensor networks for underground pipeline monitoring. Ad Hoc Netw. 2011,
9, 218–227. [CrossRef]
9. Li, H.N.; Li, D.S.; Song, G.B. Recent applications of fiber optic sensors to health monitoring in civil
engineering. Eng. Struct. 2004, 26, 1647–1657. [CrossRef]
10. Grosswig, S.; Graupner, A.; Hurtig, E.; Kühn, K.; Trostel, A. Distributed fibre optical temperature sensing
technique—A variable tool for monitoring tasks. In Proceedings of the 8th International Symposium on
Temperature and Thermal Measurements in Industry and Science, Berlin, Germany, 19–21 June 2001.
11. Niklès, M.; Vogel, B.; Briffod, F.; Grosswig, S.; Sauser, F.; Luebbecke, S.; Pfeiffer, T. Leakage detection using
fiber optics distributed temperature monitoring. Smart Struct. Mater. 2004, 18–25. [CrossRef]
12. Zou, L.; Sezerman, O.; Revie, W. Pipeline corrosion monitoring by fiber optic distributed strain and
temperature sensors. In Proceedings of the NACE International Corrosion Conference and Expo,
New Orleans, LA, USA, 16–20 March 2008.
13. Alhandawi, K.; Vahdati, N.; Shiryayev, O.; Lawand, L. Corrosion monitoring along infrastructures using
distributed fiber optic sensing. In Proceedings of the SPIE Smart Structures and Materials + Nondestructive
Evaluation and Health Monitoring, Las Vegas, NV, USA, 20–24 March 2016.
14. Li, W.; Bao, X. High Spatial Resolution Distributed Fiber Optic Technique for Strain and Temperature
Measurements in Concrete Structures. In Proceedings of the International Workshop on Smart Materials,
Structures NDT in Canada 2013 Conference & NDT for the Energy Industry, Calgary, AB, Canada,
7–10 October 2013.
Appl. Sci. 2017, 7, 540 12 of 12

15. Li, H.; Yi, T.; Ren, L.; Li, D.; Huo, L. Reviews on innovations and applications in structural health monitoring
for infrastructures. Struct. Monit. Maint. 2014, 1, 1–45. [CrossRef]
16. Ren, L.; Jia, Z.; Li, H.; Song, G. Design and experimental study on FBG hoop-strain sensor in pipeline
monitoring. Opt. Fiber Technol. 2014, 20, 15–23. [CrossRef]
17. Jia, Z.; Ren, L.; Li, H.; Ho, S.; Song, G. Experimental study of pipeline leak detection based on hoop strain
measurement. Struct. Control Health Monit. 2015, 22, 799–812. [CrossRef]
18. Felli, F.; Paolozzi, A.; Vendittozzi, C.; Paris, C.; Asanuma, H. Use of FBG sensors for health monitoring of
pipelines. In Proceedings of the SPIE Smart Structures and Materials + Nondestructive Evaluation and
Health Monitoring, Las Vegas, NV, USA, 20–24 March 2016.
19. Zhou, J.; Sun, L.; Li, H. Study on dynamic response measurement of the submarine pipeline by full-term
FBG sensors. Sci. World J. 2014, 2014. [CrossRef] [PubMed]
20. Lee, H.; Sohn, H. Damage detection for pipeline structures using optic-based active sensing. Smart Struct. Syst.
2012, 9, 461–472. [CrossRef]
21. Hou, Q.; Jiao, W.; Zhan, S.; Ren, L.; Jia, Z. Natural gas pipeline leakage detection based on FBG strain sensor.
In Proceedings of the Fifth International Conference on Measurements and Mechatronics Automation,
Hong Kong, China, 16–17 January 2013.

© 2017 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).