Process Safety and Environmental Protection 149 (2021) 894–904

Contents lists available at ScienceDirect

Process Safety and Environmental Protection

journal homepage: www.elsevier.com/locate/psep

A programmable logic controller based remote pipeline monitoring

system
Nauman Aziz a , Shujaat Ali Khan Tanoli a , Faiza Nawaz b,∗
a
COMSATS University Islamabad, Pakistan
b
University Of Alberta, Faculty of Sciences, Canada

a r t i c l e i n f o a b s t r a c t

Article history: Pipelines are vital method for long distance transportation and they need to satisfy levels of safety, unwa-
Received 20 May 2020 vering quality and efficiency. Large amount of natural resources is wasted due to leakages in pipelines.
Received in revised form 24 March 2021 In recent years, this issue has gained a lot of consideration in research community due to associated
Accepted 25 March 2021
economic losses and environmental hazards. The precise effort of this research is to design a novel leak
Available online 31 March 2021
detection system with improved sensitivity, reduce false alarm rate and higher leak localization accuracy.
The proposed test bed is established by using specific purpose sensors, A programmable logic controller
Keywords:
(PLC) and supervisory control and data acquisition (SCADA). The well-known Volume Based Method
Pipeline monitoring
Leak detection sensitivity
(VBM) and Pressure Point Analysis (PPA)approach has been adapted to propose a Leak Detection Sys-
False alarm rate tem (LSD), with improved detection sensitivity and reliability. The algorithms are deployed in field on
Leak localization accuracy pipeline test bed and performance results are documented for different testing scenarios. Results show
PLC and SCADA that the hybrid technique has leak detection sensitivity of 5 L/min in real time and ±8.5% leak localization
accuracy.
© 2021 Institution of Chemical Engineers. Published by Elsevier B.V. All rights reserved.

1. Introduction the pipeline framework (Sadeghioon et al., 2014). Pipeline leakages

can cause environmental hazards, financial loss and loss of lives.
Pipeline systems are the most economic and most secure Development in instrumentation for effective leakage detection
method of transportation for oil, gases and other liquid items. Many have been significant since 1990’s. Regardless of these advances,
countries rely upon 1000’s of km of pipeline to transport water most liquid dispersion systems continue to encounter high level
and oil from desalination plants to desired location (BenSaleh et al., of losses (Lambert, 2002). A recent United Nations Development
2013). In year 2007, Russia sent out approximately 1.3 million bar- Program (UNDP) report communicates that there has been a total
rels of raw petroleum every day by means of pipelines to Germany, of 6,817 oil slicks from 1976 to 2001, which represents a loss of
Poland and Central and Eastern Europe (Sun et al., 2011). Every about 3,000,000 barrels of oil, of which over 70 % was not recov-
year, a large amount of water and oil is wasted due to leakages in ered. Damage, corrosion, geological disaster and robbery through
pipelines, almost 60 % of water is squandered each year because oil siphoning has turned into a noteworthy issue (Foorginezhad
of leaks in pipelines (Karray et al., 2016). As per World Bank’s esti- et al., 2020). Pipeline spills can also cause loss of lives for example
mate, water leakage volume sums to 48.6 billion m3 /year around on Nov. 1, 2007, a 12-inch-width pipeline section, transporting fluid
the world, with a monetary loss of roughly 14.6 billion US dollars propane, busted in a rustic region close to Carmichael, Mississippi.
for each year (Rashid et al., 2014). The coming about fireball was heard and seen from miles away.
As a method for long distance transportation, pipelines must Around 430,626 gallons of propane were discharged. Because of
satisfy certain levels of safety, unwavering quality and efficiency. the resulting fire, two individuals were killed and seven individuals
Monitoring long distance pipelines is a difficult task because of the supported minor wounds. Large leaks have a major effect on both
inconveniences involved in keeping up the framework. Extensive pressure and flow rate, so they are easily detectable. However, for
research efforts have been made to design a vigorous and reliable smaller leaks, a better leak detection system is required. Numer-
method of monitoring spills, blasts and different irregularities in ous attempts to improve the leak detection and localization has
been made with some achievement. However, this issue is unpre-
dictable and presents numerous difficulties. These difficulties are
the inspiration for this postulation.
∗ Corresponding author.
Most of the pipelines are operated at high pressure and tempera-
E-mail addresses: [email protected] (N. Aziz), [email protected]
(S.A.K. Tanoli), [email protected] (F. Nawaz). ture, which can often result in pipelines breakdown, causing severe

https://doi.org/10.1016/j.psep.2021.03.045
0957-5820/© 2021 Institution of Chemical Engineers. Published by Elsevier B.V. All rights reserved.
N. Aziz et al. Process Safety and Environmental Protection 149 (2021) 894–904

Fig. 1. Mobile and static pipeline monitoring techniques.

damage to humans and unwanted interruption of oil, gas or water a pipeline to inspect the interior/exterior conditions of pipes.
supplies. Hence there is a need for active monitoring and main- Standalone inspection systems utilize phased array ultrasonic tech-
tenance system for these pipelines, however, the system should nology and are mounted on the outer surface of pipelines they use
remain scalable and cost effective. Pipeline inspection and repair electromagnetic waves to detect corrosion and defects at a given
presents a number of challenges that are quite unique as these point in time and space. Hand held devices are also used for manual
pipelines are often very long and they may pass through com- operation at certain location. Limited research has been conducted
pressor stations that help to maintain the required high pressure. on using a distributed network system of sensors, deployed along
Hence pipelines monitoring faces a number of challenges during its pipelines, for inspection and monitoring.
transportation these include Usually Gas pipelines are located underground or under water
therefore wired networks can result in a number of problems as
• Gas leakage, that creates the danger of fires and explosions. In discussed in (Shi et al., 2021). Once a monitoring system is deployed
addition, leaks bring about large economic losses. it becomes difficult to physically access the structure from time to
• Pipe Corrosion problem time, resulting in a demand to monitor and maintain the system
• Cracks due to aging of pipes or soil movement. remotely as much as possible. Remote monitoring can be classified
• Third-party damage into mobile and static monitoring as described in Fig. 1
• Black powder is a solid contamination that may be wet and have a Detection of minor leak and its localization requires extensive
tar-like appearance. It can cause contamination, erosion and flow study, projection and planning. Various smart methods are emerg-
reduction. Chemical analyses of the material have revealed that ing to remotely monitor various pipelines. IoT-based solutions are
it consists mainly of a mixture of and iron sulfides. gaining popularity as they can collect data is Realtime with high
accuracy and efficiency (Abdelhafidh et al., 2020). These systems
To overcome these challenges several solutions have been are secure and offer resilience against various types of attacks. A
proposed over years. Pipeline Monitoring system can be wired, recent study (Sajid et al., 2016) presents the security challenges and
wireless or a combination of both to transfer data from pipelines solutions that are linked with SCADA and IoT based pipeline moni-
to control station. Since pipeline networks are in line, connectivity toring systems. The research suggests that by integrating complex
always remains a critical feature. Pipeline inspection can be broadly IoT with wireless sensor network and cloud computing, various
classified into three major categories monitoring issues of classical systems can be resolved.
Rest of the paper is organized as described next. Section 2 pro-
• Using intelligent mobile sensing devices, vides a background of pipeline monitoring techniques, highlighting
• Using standalone inspection systems, instruments the major contributions within this field. In this section Hardware
• Using resident pipeline sensors. based and software-based leak detection systems are discussed
and compared. Section 3 presents the research process adopted
for implementing our proposed leak detection system design and
Intelligent mobile sensing devices are robots, pigs, or crawlers
implementation methodology. Testbed design, hardware compo-
these are intelligent robotic devices that can move inside or above

895
N. Aziz et al. Process Safety and Environmental Protection 149 (2021) 894–904

Fig. 2. Types of Pipeline leak detection systems.

nents and mathematical model is presented in this section. Section The integrity of pipelines can also be monitored utilizing assess-
4 presents the results and list some future research directions. ment devices like intelligent pigs or smart robots. The material
condition of a pipeline is surveyed by determining magnetic flux
2. Background and related work leakage (MFL). The MFL strategy is performed by polarizing the steel
pipe near to the saturation flux density and after that recognizing a
There are various perspectives to classify Leak Detection Sys- nearby flux leakage caused by surface irregularities. This method-
tems (LDS). Baroudi et al. (2019) have classified them into various ology gives an exact estimation of wall thickness only at point of
categories and study their data fusion techniques. Most popular physical contact and not along the boundary of the pipe (Ege and
LDS systems are named as biological methods (leak detection by Coramik, 2018).
visual inspection, odor or sound), hardware Based methods and
software-based methods. Depending on the development environ- 2.2. Software based methods
ment, Abdelhafidh et al. (2018) divides LDS systems into three
classes, named as above ground pipeline monitoring, underground In Software based pipeline monitoring methods, the sensors are
pipeline monitoring and under water pipeline monitoring. Authors placed inside the pipeline in order to measure the internal param-
in (Wang and Ye, 2010) categorized pipeline monitoring into two eters such as viscosity, temperature, flow rate, pressure and infer a
types, named as noninvasive techniques where sensors are placed commodity release by computation. Mass/Volume Balance method
outside pipeline and invasive techniques where sensors are placed is often used for accurate leak detection in long distance pipeline.
inside pipeline. A broad classification of Pipeline monitoring sys- In (Martins and Seleghim, 2010) a novel technique is introduced in
tems (PMS) is shown in Fig. 2. which if there is a contrast between the upstream and downstream
stream in excess of a specific limit then a leak caution is produced.
2.1. Hardware based methods This technique is known as ‘mass balance method’. The leak can
be detected by using flow sensors inside the pipeline. The advan-
In this monitoring technique the sensors are placed outside tage of this system is that it can detect very small leaks and it is
the pipeline to perform monitoring and reporting. This technique cost efficient. The disadvantage of this technique is that it cannot
normally operates on the non - algorithmic principal of physi- precisely localize the position of the leakage and the rate of false
cal detection of escaping commodity. One of the most popular alarms is more frequent because of many factors such as block-
externally based methods is Soil Monitoring. In this procedure, the age of fluid inside the pipeline. Negative Pressure Wave (NPW) is
pipeline is first implanted with a little measure of a tracer com- another promising leak detection technique. Whenever a leak is
pound, which holes out of the pipe if there is any occurrence of developed with in a pipeline the density of the liquid or gas at that
an opening. This is perceived by hauling an instrument along the point decreases and from that leak point a wave equal in magnitude
surface over the pipeline (Aalsalem et al., 2018). The advantage of but opposite in direction propagates through the pipeline. This phe-
this sort of strategy is that it incorporates low false cautions and nomenon is known as negative pressure wave (Wan et al., 2012).
is high affectability Distributed fiber optical cable sensing has a In Pressure Point Analysis (PPA) the pressure inside the pipeline
unique feature to determine temperature at multiple points for is monitored. If it drops frequently then there is a spillage. Pres-
monitoring long distance pipelines (Ren et al., 2018). The ultra- sure at various focuses along the pipeline is consistently observed
sonic testing technique is also a promising solution that utilizes utilizing factual investigation of these estimations, the nearness of
high frequency mechanical vibrations to decide the wall thickness hole is proclaimed when mean estimation of pressure diminishes
of a pipe with high precision and unwavering quality (Chen et al., under a predefined limit (Abdulshaheed et al., 2017). An infer about
2020). It gives imperfections indication when there is variation leakage event can also be made on the basis of statistical analysis
in the pipeline wall thickness. This system enables a light beam of typical field data such as flow rate, pressure and temperature.
emission of ultrasonic energy over the pipe surface. The technique Real Time Transient Model (RTTM) can simulate pipeline condi-
includes setting the sensor probes at a specific point to the pipeline tions using the hydraulic modeling and advance fluid modeling.
wall. The probes send high frequency ultrasonic energy which is RTTM software typically uses the conservation of energy equation,
transmitted through the pipe as waves. Some portion of the wave conservation of momentum, and various flow equations to mea-
is reflected in reverse from where there is irregularity. The wave sure the fluid flow. This method can predict the size and location
speed and time of travel would then be able to evaluate the distance of leaks in real time by comparing the calculated values with the
of flaw in the pipe walls (Xianming et al., 2020). actual data. It is possible to calculate real time pipeline parame-

896
N. Aziz et al. Process Safety and Environmental Protection 149 (2021) 894–904

ters such as flow, pressure and temperature using modern digital with application in water pipeline monitoring. The leak detection
computers. This is known as RTTM system of leakage detection and sensitivity of this system is approximately 5−60 L/min and it can
localization (Okungbowa and Brown, 2018). also localize the leaks with an accuracy of ∼ 5 feet in 0.1 s.

2.3. Limitations of existing leak detection systems 3.1. Remote pipeline monitoring system design

There are limited instruments available for pipeline leak detec- PLC takes a feedback signal through field input devices and then
tion and every hardware-based procedure has some limitations. constantly monitors the state of the system with a control applica-
The choice of a hardware-based method is ought to precisely detect tion program that is executed with in it. SIEMENS S7−1200 family
and localize the spills during steady and transient state. Acoustic member 1214c PLC is used in this research project.
sensors have limited detection range. Magnetic flux leakage (MFL) There are three options available for the S7−1200 series i.e.
accuracy decreases as the thickness of material increases and MFL 1211c, 1212c and 1214c. Among these three, the 1214c has been
is limited to only magnetic materials with eddy current method. used in this research because of the following three characteristics.
Fiber optic system have the disadvantages of not being scalable,
any damage to the sensing cable can render the whole framework, 1 It has a compact high-performance design with 24 integral
making it non-practical. Installation cost of this method is also high input/outputs, expandable by 1 signal board and 8 signal mod-
as cable is to be installed along the pipeline (Tsukada et al., 2017). ules.
In contrast to the hardware technique, the software leak 2 It comes with a 24 V encoder that helps in establishing direct
detection method utilizes instrumentation to quantify internal connection of sensors and encoders.
parameters of the pipeline. The more parameters that are utilized 3 Comes with a simulator for simulating the integrated inputs and
for a specific software strategy, the more accurate the outcomes will for testing the user program.
be. The limitations of software-based leak detection methods are
as follows: Volume balancing and Pressure point analysis have high Sensing unit collects the data from distributed input output sta-
false alarm rate and their performance is limited in some scenarios. tion. In this work, FS-400A-G1 flow meter sensor and HK1100c
RTTM method is difficult to maintain and also have false alarm rate. pressure sensors have been utilized. The distributed input out-
Leak localization accuracy of statistical analysis method is not up to put station diagram of setup is shown in Fig. 2. Communication
the mark. Negative Pressure Wave method requires dedicated leak between control center and remote monitoring station is estab-
detection hardware (Liu et al., 2017; Obeid et al., 2016). lished using wireless radio transceiver as it is versatile, efficient and
stable customer premises equipment (CPE). Radio is equipped with
a high output power of up to 28 dB and its receiving sensitivity vary
3. Methods
between -97 and -75 dB. SCADA enables the operator to perform
the tasks remotely. It provides accurate information in real time, it
In perspective of the above limitations it is quite evident that
assists the operator to control and monitor the oil and gas pipeline
tremendous efforts are needed for the advancement of an intelli- operations. PLCs take the key measurements like flow rate, pressure
gent methodology that can detect the leakage, and also localize it and temperature from the field devices and upload to host SCADA
with accuracy. This research work is an effort to present an inher- via communication network. In response to the incoming data the
ently intelligent strategy towards this direction. The requirement operator may then issue commands of supervisory nature to the
of the industry to select a LDS depends on the performance of that field location (Fig. 3).
system. American Petroleum Institute (API) provides the perfor- This leak detection model adopts the mode of distributed control
mance criteria of a LDS and this simplifies the selection of LDS for and centralized monitoring. The system is divided into three levels.
the customer. This performance criteria includes reliability, sensi- The first level is distributed field input output station. Here, fluid
tivity, stability and accuracy. Programmable logic controller (PLC) pump, sensors and control valves are erected along the pipeline.
and Supervisory Control and Data Acquisition (SCADA) based LDS At this level various sensor types can be used to monitor dif-
that focuses in improving these performance matrices and imple- ferent parameters. The second station is master control station
mented for oil pipeline is given in (Obeid et al., 2016). Proposed which is responsible for data acquisition from field devices and
system provides a leak localization accuracy of 120 m in 0.1 s and to successfully implement the proposed algorithms. It also aids in
leakage detection sensitivity with transportation is approximately transmitting the data to remote monitoring station. Here we can
10−40 L/min. This research work also focuses to improve the per- test various radios to improve and test the performance of data
formance matrices such as the reliability of leakage detection, the delivery under various scenarios. The third level is remote moni-
leakage detection sensitivity and the leakage localization accu- toring station, whose functions are to control pump/valves, check
racy by employing advance integrated techniques such as volume pipeline operating state and to display graphical results and engi-
balancing method, pressure point analysis method and negative neering variables like flow rate. The detailed system model of this
pressure wave method. architecture is shown in Fig. 4.
The main contributions of this research work are as follows:
3.2. Testbed and hardware setup
1 Design of an efficient leak detection system for liquid transporta-
tion. To design an efficient leak detection system test-bed, several
2 System design focuses on the improvement of leakage detection parameters were considered including the pipe material, its size,
reliability, leakage detection sensitivity and leakage localization pump capacity, the reservoir size, the setup estimate, valve struc-
accuracy. ture and its design. The listed parameters have been examined and
3 The proposed design is integrated with remote monitoring sta- a laboratory framework setup have been designed in our research
tion (RMS). lab so as to test the PLC and SCADA based LDS. Fig. 5 shows the
test-bed of pipeline monitoring system, composed of approx. 22
This research work deals with the physical designing of a PLC feet long 32 mm external diameter of polyethylene pipe. The setup
and SCADA based pipeline monitoring system to share experience comprises of three valves in order to fluctuate the pressure, flow
with research community and to insight in implementing a test-bed rate and to induce leakages in pipeline. A 2000-liter reservoir is uti-

897
N. Aziz et al. Process Safety and Environmental Protection 149 (2021) 894–904

Fig. 3. Distributed input output station diagram.

Fig. 4. System model of the proposed leak detection system.

lized as a water source. To measure the upstream and downstream occurs. The difference between the in and out product volumes in
flow rate four flow meter sensors are utilized. Pressure sensors are a particular segment of the pipeline is written as (1)
installed to measure internal pressure at both ends of pipeline. An  
electric pump with 1 hp motor is employed to provide up to 6 bars  Qin − Qout  ≤ dQm + dVs (1)
t
of pressure.
Where
Qin = Measure of fluid flow in
Qout = Measure of fluid flow out
3.3. Mathematical modelling for leak detection dQm = Bound of uncertainty in flow measurement
dVs
= Bound of uncertainty in pipeline inventory over a time intervalt.
t
Leakage in pipeline changes the flow rate at inlet and outlet. If leak exist in the pipeline leak flow rate Ql can be detected
Also, a sudden leak can result in abrupt pressure change in pipeline. when the following relation is fulfilled.
This proposed system uses both of these phenomena in order to
  dVs
increase the leak detection reliability and set a threshold value to Ql =  Qin − Qout  > dQm + (2)
improve leak detection sensitivity. The proposed technique applies t
the Volume Based Method (VBM) and Pressure Point Analysis (PPA)
method to detect pipeline leaks and then Negative Pressure Wave Pressure sensors mostly are introduced at the boundaries of
(NPW) method is utilized to locate the exact location where leak pipelines. At the point when the Pressure has achieved enduring

898
N. Aziz et al. Process Safety and Environmental Protection 149 (2021) 894–904

3.4. Mathematical modelling for leak localization

When a leak is developed in a pipeline the density of the liquid

or gas at that point will decrease and from that leak point a wave
equal in magnitude but opposite in direction propagates through
the pipeline (Tian et al., 2012). This phenomenon is known as neg-
ative pressure wave. The pressure sensors installed upstream and
downstream from that leak point collects the negative pressure
wave signals. From the time difference for detecting the exact leak-
age point can be detected (Lu et al., 2006). This principle is shown
in the Fig. 6.
The distance between the leak point and upstream sensor can
be obtained by Eq. (5).
x
t1 =
v−u
L−x (5)
t2 =
v+u
t = t1 − t2

The distance between the leak point and upstream sensor can
be obtained by Eq. (6) (Rashid et al., 2014).
L + v.t
x= (6)
2
In Eq. (6)
x= the distance between leak point and pressure transducer.
L= the distance between two pressure transducers
v= negative pressure wave propagation velocity in liquid
medium m/s.
t= time difference of pressure wave getting to both pressure
transducers on pipe.
All the valves (V-1, V-2 and V-3) were closed before the com-
mencements of experiments. Motor was set to produce a transient
flow and was kept about for 1 min until transient flow approached
steady state flow. Pipeline flow rate and internal pressure were
Fig. 5. a) 22 feet long, 32 mm external diameter polyethylene pipe. The setup com-
prises of three valves in order to fluctuate the pressure, flow rate and to induce measured during no leak condition. After the steady state flow, the
leakages in pipeline. 5b) Master control station consisting of data processing unit. valve V-1 opening was kept open, then valve was regulated to dis-
tinguish different leak rates. The change in flow rate, the leakage
flow rate, and the pressure drop and leakage localization values
state, a limit is set with the end goal that is the Pressure falls under- were measured. Then percentage difference between actual and
neath a pipeline there is likely a leak in the pipeline. Pressure drop calculated leakage location was measured. Similarly, readings were
between inlet and outlet can be calculated using equation below taken for valves V-2 and V-3.
(Chuka et al., 2021).
4. Results and discussion
P = PI − PO (3)
This section is arranged as follows; firstly, leak detection tests
Here, P is change in pressure, PI pressure at inlet and PO pres- are conducted using VBM and PPA method and setting up the
sure at outlet. The pressure drop subsequently can be compared threshold values to evaluate reliability and sensitivity of leak detec-
against a threshold A to evaluate a leakage Alarm. tion. Secondly; leak localization tests are conducted using NPW
  method by generating leaks at three different locations and cal-
dp p ≤ −A ⇒ No Leak culating measured verses actual leak location distances to evaluate
≈ (4)
dt t > −A ⇒ Leak leak localization accuracy.

Fig. 6. Negative pressure wave phenomenon.

899
N. Aziz et al. Process Safety and Environmental Protection 149 (2021) 894–904

Fig. 7. Flow rate of FS 1, FS 2, FS 3 and FS 4with all valves closed.

Fig. 8. a) Inlet and outlet flow rates with open valve V1 8b) Magnitude of leakage with open valve V1.

Fig. 9. a) Inlet flow sensors measurement with valve V2 opened, 9b) Outlet Flow sensors measurement with valve V2 opened.

900
N. Aziz et al. Process Safety and Environmental Protection 149 (2021) 894–904

Fig. 10. Leak Detection using PPA. a) Pressure drop due to Leak 1, b) Pressure drop due to Leak 2, c) Pressure drop due to Leak 3.

901
N. Aziz et al. Process Safety and Environmental Protection 149 (2021) 894–904

Fig. 11. Leak Point Localization using NPW a) leak 1 location, b) leak 2 location, c) leak 3 location.

4.1. No leak condition 4.2. Leak detection using VBM

To check normal flow rates of flow meters, all valves were turn It has been observed that there is often a sudden increase in
closed and flow rates were measured. Graphical results of FS 1, FS inlet flow as there is a decrease in pipe flow resistance between
2, FS 3 and FS 4 are shown in Fig. 7. upstream flow meter and the leak. While, there is an abrupt fall in

902
N. Aziz et al. Process Safety and Environmental Protection 149 (2021) 894–904

the outlet flow because the fluid leaves through the leakage point localization accuracy of the system is ±8.5 % in 0.1 s. the proposed
instead of passing through the downstream flow sensor. In our sys- design can be extended by implementing the same LDS using wire-
tem, leakage was generated by opening valve V-1. Flow rates of flow less sensors. This will increase the reliability of the LDS and reduce
sensors deviates from their normal point operation. The location of the wire maintenance cost.
valve V-1 is between FS 1 and FS 2. Fig. 8 shows that as the leak is
generated the flow rate of inlet flow sensor increases from 24 L/min
Declaration of Competing Interest
to 25.25 L/min and flow rates of outlet flow sensor decreases from
20 L/min to 19 L/min. Fig. 7b shows magnitude of leakage between
The authors report no declarations of interest.
FS1 and FS2 that is approximately 7 L/min. It also indicates the leak
duration (Fig. 9).
Next, leak is generated by opening valve V-2 which is located Acknowledgement
between FS2 and FS3. Change in flow rates of inlet flow sensors
FS1, FS2 and outlet flow sensors FS3, FS4 were measured. Fig. 10a) We would like to thank the Department of Electrical and Com-
shows graphical results where it is clear that the leak detection puter Engineering, COMSATS University Islamabad, Attock Campus,
condition is fulfilled, that is the flow rate of FS located after the leak for providing the favorable research environment. We are also
point decreases and flow rate of FS before the leak point increases. obliged to Higher Education Commission (HEC) for providing
A 6 L/min magnitude of leak was detected in this setting. National Research Program for Universities (NRPU) 5306/Fed-
Similarly, when valve V-3 was set in open position, which is eral/NRPU/R&D/HEC/2016 funding to support this research work.
located between flow sensor 3 and flow sensor 4, there was a sud-
den increase in inlet flow rate and a sudden decrease in outlet flow
rate. References

Aalsalem, M.Y., Khan, W.Z., Gharibi, W., Khan, M.K., Arshad, Q., 2018. Wireless sensor
4.3. Leak detection using pressure point analysis networks in oil and gas industry: recent advances, taxonomy, requirements, and
open challenges. J. Netw. Comput. Appl. 113, 87–97.
Pressure drop occur inside the pipeline whenever a leak is Abdelhafidh, M., Fourati, M., Fourati, L.C., Laabidi, A., 2018. An investigation on wire-
less sensor networks pipeline monitoring system. Int. J. Wirel. Mob. Comput. 14,
present. The proposed PPA method is based on the relative change
25–46.
in pressure profile in the pipe. After calibration of pressure trans- Abdelhafidh, M., Fourati, M., Chaari Fourati, L., 2020. A genetic algorithm-based
mitter, it gives reading from 0 to 1.2 MPa. The laboratory tests are intelligent solution for water pipeline monitoring system in a transient state.
Concurr. Comput. Pract. Exp., e5959.
illustrated in Figures bellow. There are four main phases of this
Abdulshaheed, A., Mustapha, F., Ghavamian, A., 2017. A pressure-based method for
experiment that are pump start, stabilization of pressure readings, monitoring leaks in a pipe distribution system: a review. Renew. Sustain. Energy
sudden pressure drops due to leakage and after some time leak- Rev. 69, 902–911.
age is turned off. In pressure stabilization phase, the pressure at Baroudi, U., Al-Roubaiey, A.A., Devendiran, A., 2019. Pipeline leak detection systems
and data fusion: a survey. IEEE Access 7, 97426–97439.
both ends of pressure transmitter is fluctuating between 116 K Pa BenSaleh, M.S., Qasim, S.M., Obeid, A.M., Garcia-Ortiz, A., 2013. In review on wireless
to 116.5 K Pa. Then leakage was generated by opening valve V- sensor network for water pipeline monitoring applications. In: 2013 Interna-
1, pressure of PT-1 drops to 114.5 K Pa and value of PT-2 drops to tional Conference on Collaboration Technologies and Systems (CTS), IEEE, pp.
128–131.
115.1 K Pa. This drop clearly indicates leakage in pipeline. Similarly, Chen, Y., Chen, Y., Hu, S., Ni, Z., 2020. Continuous ultrasonic flow measurement for
this experiment was repeated for valves V-2 and V-3 and change in aerospace small pipelines. Ultrasonics 109, 106260.
pressure profile was noted. Chuka, C.E., Freedom, I., Anthony, O., 2021. Transient Model-based Leak Detection
and Localization Technique for Crude Oil Pipelines: a Case of Npdc, Olomoro.
CV-Zhifeng Zhang.
4.4. Leak point localization using negative pressure wave Ege, Y., Coramik, M., 2018. A new measurement system using magnetic flux leakage
method in pipeline inspection. Measurement 123, 163–174.
Foorginezhad, S., Mohseni-Dargah, M., Firoozirad, K., Aryai, V., Razmjou, A., Abbassi,
When a leak is developed in a pipeline the density of the liq-
R., Garaniya, V., Beheshti, A., Asadnia, M., 2020. Recent advances in sensing and
uid or gas at that point will decrease. From that leak point a wave assessment of corrosion in sewage pipelines. Process. Saf. Environ. Prot.
equal in magnitude but opposite in direction propagates through Karray, F., Garcia-Ortiz, A., Jmal, M.W., Obeid, A.M., Abid, M., 2016. Earnpipe: A
the pipeline. This phenomenon is known as negative pressure wave. testbed for smart water pipeline monitoring using wireless sensor network.
Procedia Comput. Sci. 96, 285–294.
Eq. (5) is used to find the leakage locations of leak points V-1, V-2 Lambert, A.O., 2002. International report: Water losses management and techniques.
and V-3. The graphical results of leakage points are shown in Fig. 11. Water Sci. Technol. Water Supply 2, 1–20.
Actual and calculated results were compared and percentage Liu, C., Li, Y., Fang, L., Han, J., Xu, M., 2017. Leakage monitoring research and design
for natural gas pipelines based on dynamic pressure waves. J. Process Control
error was found. It is observed that smaller the values relative error, 50, 66–76.
larger the localization accuracy of the LDS. For leak point V1 the Lu, X., Sang, Y., Zhang, J., Fan, Y., 2006. In pipeline leakage detection technology
actual distance was 4.6 ft and calculated distance was 4.8 ft, giving based on wavelet transform theory. In: 2006 IEEE International Conference on
Information Acquisition, IEEE, pp. 1432–1437.
a relative error or 4.34 %. For valve V2 actual distance was 14 ft Martins, J.C., Seleghim, P., 2010. Assessment of the performance of acoustic and mass
obtained results showed the distance as 12.8 ft. similarly for valve balance methods for leak detection in pipelines for transporting liquids. J. Fluids
V3 its actual position was 10 ft and measured value obtained was Eng., 132.
Obeid, A.M., Karray, F., Jmal, M.W., Abid, M., Qasim, S.M., BenSaleh, M.S., 2016.
18.6 ft, giving a relative error of 3.33 %. Towards realisation of wireless sensor network-based water pipeline monitor-
ing systems: A comprehensive review of techniques and platforms. IET Sci. Meas.
5. Conclusion Technol. 10, 420–426.
Okungbowa, N., Brown, T., 2018. In strategies for isothermal, real-time, transient
pipeline models in shut-in conditions. In: International Pipeline Conference,
In this research a unique LDS has been designed and tested, American Society of Mechanical Engineers, p V003T004A020.
employing the VBM and PPA as basic principle to improve leak Rashid, S., Qaisar, S., Saeed, H., Felemban, E., 2014. A method for distributed pipeline
burst and leakage detection in wireless sensor networks using transform anal-
detection sensitivity and leak detection reliability. NPW method is
ysis. Int. J. Distrib. Sens. Netw. 10, 939657.
also used in order to improve leak localization accuracy. According Ren, L., Jiang, T., Jia, Z.-g., Li, D.-s., Yuan, C.-l., Li, H.-n., 2018. Pipeline corrosion and
to results, this hybrid system has justified calculated expectations, leakage monitoring based on the distributed optical fiber sensing technology.
which indicates that the given method has achieved its technical Measurement 122, 57–65.
Sadeghioon, A.M., Metje, N., Chapman, D.N., Anthony, C.J., 2014. Smartpipes: Smart
limit. Results show that that the system is leak sensitive and can wireless sensor networks for leak detection in water pipelines. J. Sens. Actuator
detect a minimum 5 L/min leak and maximum 60 L/min. The leak Netw. 3, 64–78.

903
N. Aziz et al. Process Safety and Environmental Protection 149 (2021) 894–904

Sajid, A., Abbas, H., Saleem, K., 2016. Cloud-assisted iot-based scada systems secu- Tsukada, K., Majima, Y., Nakamura, Y., Yasugi, T., Song, N., Sakai, K., Kiwa, T., 2017.
rity: a review of the state of the art and future challenges. IEEE Access 4, Detection of inner cracks in thick steel plates using unsaturated ac magnetic flux
1375–1384. leakage testing with a magnetic resistance gradiometer. IEEE Trans. Magn. 53,
Shi, G., Yu, W., Wang, K., Dang, F., Gong, J., Lu, Y., 2021. Time-dependent economic risk 1–5.
analysis of the natural gas transmission pipeline system. Process. Saf. Environ. Wan, J., Yu, Y., Wu, Y., Feng, R., Yu, N., 2012. Hierarchical leak detection and local-
Prot. 146, 432–440. ization method in natural gas pipeline monitoring sensor networks. Sensors 12,
Sun, Z., Wang, P., Vuran, M.C., Al-Rodhaan, M.A., Al-Dhelaan, A.M., Akyildiz, I.F., 189–214.
2011. Mise-pipe: Magnetic induction-based wireless sensor networks for under- Wang, G., Ye, H., 2010. Leak Detection and Positioning in Fluid Transmit Pipeline.
ground pipeline monitoring. Ad Hoc. Netw. 9, 218–227. Tsinghua University Press, Beijing.
Tian, C.H., Yan, J.C., Huang, J., Wang, Y., Kim, D.-S., Yi, T., 2012. In negative pressure Xianming, L., Ping, L., Baocun, Z., Jiangtao, C., Ying, G., Zhe, K., Siyu, L., 2020. Local-
wave based pipeline leak detection: challenges and algorithms. In: Proceedings ization of multiple leaks in a fluid pipeline based on ultrasound velocity and
of 2012 IEEE International Conference on Service Operations and Logistics, and improved gwo. Process. Saf. Environ. Prot. 137, 1–7.
Informatics, IEEE, pp. 372–376.

904