UNIVERSITI TEKNOLOGI MALAYSIA
PSZ 19:16 (Pind. 1/13)
UNIVERSITI TEKNOLOGI MALAYSIA
DECLARATION OF THESIS / UNDERGRADUATE PROJECT PAPER
Author’s full name
: NURUL HIDAYAH BINTI ABDUL WAHAB
Date of birth
: 12nd APRIL 1991
Title
: FAULT LOCATION OF FAULTY UNDERGROUND CABLE IN A NOISY
ENVIRONMENT BASED ON ACOUSTIC SIGNALS
Academic Session
: 2014 / 2015
I declare that this thesis is classified as :
CONFIDENTIAL
(Containing confidential information under the Official Secret
Act 1972)*
RESTRICTED
(Containing restricted information as specified by the
organisation where the research was done)*
OPEN ACCESS
I agree that my thesis be published and accessed online (full
text)
I acknowledge that Universiti Teknologi Malaysia reserves the right as follows:
1. The thesis is the property of Universiti Teknologi Malaysia.
2. The Library of Universiti Teknologi Malaysia has the right to make copies for academic
purposes.
Certified by :
SIGNATURE
SIGNATURE OF SUPERVISOR
910412-06-5058
DR. ZURAIMY BIN ADZIS
(NEW IC NO. /PASSPORT NO.)
Date : 28th JUNE 2015
NOTES :
*
NAME OF SUPERVISOR
Date : 28th JUNE 2015
If the thesis is CONFIDENTAL or RESTRICTED, please attach the letter from
the organisation concerned stating the reason/s and duration for the
confidentiality or restriction.
“I declare that I have read this work and in my opinion this work is adequate in
terms of scope and quality for the awarding of the degree of
Bachelor of Engineering (Electrical)”
Signature
:
Name of Supervisor :
Date
:
DR. ZURAIMY BIN ADZIS
JUNE 2015
FAULT LOCATION OF FAULTY UNDERGROUND CABLE IN A NOISY
ENVIRONMENT BASED ON ACOUSTIC SIGNALS
NURUL HIDAYAH BINTI ABDUL WAHAB
A report submitted to the Faculty of Electrical Engineering
in partial fulfilment of the requirements for the award of
the degree of Bachelor of Engineering (Electrical)
Faculty of Electrical Engineering
Universiti Teknologi Malaysia
JUNE 2015
ii
“I declare that this undergraduate project paper entitled, “Fault Location of Faulty
Underground Cable in A Noisy Environment Based on Acoustic Signals” is the
result of my research except as cited in the references. The undergraduate project
paper has not been accepted for any degree and is not concurrently submitted in
candidature of any other degree”
Signature
:
Name
: NURUL HIDAYAH BINTI ABDUL WAHAB
Date
:
JUNE 2015
iii
To my late parents
Abdul Wahab Bin Jaafar and Norehan Binti Nasir
My siblings
Hisham, Fidzlah, Junizam, Anwar, Azlan, Aini
My friends
Kufazilah, Farhana, Nadiah, Naja’a, Arina, Dhamirah
Dedicated in thankful appreciation for your support, encouragement and best wishes
iv
ACKNOWLEDGEMENT
In the name of Allah S.W.T, the Most Beneficent and the Most Merciful. It is
deepest sense appreciation of the Almighty that give me strength and ability to
complete this report.
Foremost, I would like to express my sincere gratitude to my project
supervisor, Dr. Zuraimy Bin Adzis for his continuous offers of precious guidance
and advice during the construction of this project. Besides knowledge, the most
important thing I gained invaluable experience and skills that will be useful in
conducting further research towards becoming an excellent engineer.
I would like to give my appreciation to Ir. Dr. Mokhtar Bin Harun for his
guiding and knowledge sharing that related to this project. My heartily thanks also go
to Ms Nordiana Azlin Bt Othman who assisted me in the experimental works as well
as all technicians in Institute of High Voltage and High Current (IVAT) laboratory
for sharing their experiences with me.
My appreciation also goes to the most important persons in my life,
my beloved family and best friend who have been continuously giving me all the
support for all these years. Last but not least, thanks to all my friends and individuals
who has given the constructive comments and ideas during this project. In return, I
hope this research will give the advantages and knowledge for all readers. May Allah
reward you with guidance, health, and prosperous in life.
v
ABSTRACT
Underground cable systems are exposed to numerous hazards that lead to
insulation failure. Quick detection, locating and either replace or repair will increase
reliability of the power system. Observation from on-site fault locating technique
nearby roads shows that the audible noise from the passing motor vehicles hinders
the detection of electrical discharge audible signal. This causes the fault locating to
be done at night where less noise is present. In this project, the point electrode
configuration was taken as representation of faulty cable in order to generate the
spark discharge. The audible sound from spark was recorded using two methods, i.e.,
single microphone and double microphones. The Matlab R2010a and Praat software
were used to analyse the spark generated for each recording method. The results
from the experiment are presented in the time domain and frequency spectrum of the
spark discharge. The analysis of single microphone shows the appearance and
existence of the background noise as the microphone placed farther from the spark
location. Meanwhile, the findings from the double microphone indicate the delaying
time is increased as the gap distance between microphones increased and further
from the spark point. However, the background noise for this method is not giving
impact to the captured spark pulse. Likewise, the interruption towards the spark
generated also presented in this report to show that audible noise from environment
could be affect the identification of faulty point in an underground cable. Hence, the
improvements in detecting fault point should be done to enhance the reliability of the
power system network.
vi
ABSTRAK
Sistem kabel bawah tanah mudah terdedah kepada ancaman yang
menyebabkan kegagalan penebat. Kebolehharapan sistem kuasa mampu ditingkatkan
jika mngesanan lokasi kerosakan dengan kadar segera, lantas proses pembaikian
kabel baru boleh dijalankan dengan cepat. Tinjauan telah dijalankan di lokasi kabel
rosak, didapati bahawa proses pengecaman bunyi yang terhasil di lokasi rosak akan
mudah terpengaruh dengan bunyi yang terdapat di sekeliling seperti bunyi kereta.
Hal ini menyebabkan proses pengesanan kerosakan perlu dilakukan pada waktu
malam demi mengelak dari bunyi persekitaran. Menerusi projek ini, konfigurasi poin
elektrod telah digunakan bagi mewakili kabel yang rosak bagi menjana bunyi
percikan. Bunyi percikan yang terhasil telah dirakam dengan menggunakan dua
kaedah, iaitu, penggunaan satu mikrofon dan dua mikrofon. Perisian Matlab R2010a
dan Praat telah digunakan untuk menganalisis bunyi percikan bagi setiap kaedah
rakaman. Hasil eksperimen diterjemahkan dalam bentuk domain masa dan frekuensi
spektrum. Analisis daripada penggunaan satu mikrofon menunjukkan kewujudan
bunyi luar apabila mikrofon diletakkan berjauhan dari titik percikan. Manakala, hasil
eksperimen yang menggunakan dua mikrofon menunjukkan bahawa terdapat
perbezaan masa antara dua gelombang yang telah dirakam secara serentak. Tetapi,
bunyi sekeliling tidak memberi kesan yang ketara kepada gelombang percikan.
Selain itu, gangguan terhadap bunyi percikan ditunjukkan dalam laporan ini bagi
membuktikan bahawa bunyi luar mampu mengganggu bunyi percikan yang terhasil
di titik kabel rosak. Oleh itu, peningkatan dalam mengesan titik kesalahan yang perlu
dilakukan untuk meningkatkan kebolehpercayaan rangkaian sistem kuasa.
vii
TABLE OF CONTENTS
CHAPTER
TITLE
PAGE
DECLARATION
ii
DEDICATION
iii
ACKNOWLEDGEMENT
iv
ABSTRACT
v
ABSTRAK
vi
TABLE OF CONTENTS
vii
LIST OF FIGURES
x
LIST OF TABLES
xiii
LIST OF SYMBOLS AND ABBREVIATIONS xiv
1
2
INTRODUCTION
1
1.1
Overview of the project
1
1.2
Problem statement
2
1.3
Objectives of the project
3
1.4
Scope of the project
3
1.5
Report structure
4
1.6
Gantt’s chart
6
LITERATURE REVIEW
8
2.1
Introduction
8
2.2
Definitions of cable fault
9
2.3
Causes of cable fault
9
viii
3
4
2.4
Types of cable fault
10
2.5
Circuit of cable fault
11
2.6
Procedure for cable fault location
12
2.7
Pinpointing activity
13
2.8
Acoustic method
14
2.9
Audible sound
16
2.10
Sound spectrum
17
2.11
Fast Fourier Transform
17
2.12
Stereophonic and binaural system
18
2.13
Summary
20
METHODOLOGY
21
3.1
Introduction
21
3.2
Project flow
21
3.3
Equipments
23
3.4
Generation of spark discharge
29
3.5
Placement of microphones
29
3.5.1
Single microphone recording
30
3.5.2
Double microphone recording
31
3.6
Experimental setup
32
3.7
Wave file analysis
33
3.8
Summary
34
RESULTS AND DISCUSSION
35
4.1
Introduction
35
4.2
Single microphone recording results
35
4.2.1
Time domain waveform of spark
sound
35
4.2.2
Intensity level of spark sound
38
4.2.3
Frequency spectrum of spark sound 39
ix
4.3
Double microphone recording results
4.3.1
4.4
4.5
5
40
Time domain waveform of spark
sound
40
4.3.2
Intensity level of spark sound
43
4.3.3
Frequency spectrum of spark sound 44
Spark sound with interference
45
4.4.1
Human interference
45
4.4.2
Passing car’s sound interference
46
Summary
47
CONCLUSION AND RECOMMENDATIONS 48
5.1
Conclusion
48
5.2
Recommendations
51
REFERENCES
52
APPENDIX A
55
x
LIST OF FIGURES
FIGURE NO.
TITLE
PAGE
2.1
Equivalent circuit of cable fault
12
2.2
Acoustic sound produced in faulty cable
14
2.3
Pinpointing equipment by BAUR company
15
2.4
Stereophonic system (two pickups and two loudspeakers)
19
2.5
Binaural system (two pickups and two headphones)
19
2.6
Diotic system (one pickup and two headphones)
20
3.1
Project flow
22
3.2
High Voltage transformer
24
3.3
Voltage regulator
24
3.4
Diode
25
3.5
Capacitor
25
3.6
Insulator
26
3.7
Voltage divider
26
3.8
Measuring spark gap and point electrodes
27
3.9
Discharge rod
27
xi
3.10
Microphones
28
3.11
Point electrodes configuration in spark generation
29
3.12
Position of single microphone
30
3.13
Experiment of single microphone recording
30
3.14
Position of double microphone
31
3.15
Experiment of double microphone recording
31
3.16
Experimental circuit in IVAT laboratory
32
3.17
Experimental setup of double microphone recording
33
4.1
Time-domain waveform at various distances, D
36
4.2
Time taken for the first spark pulse
37
4.3
Sound intensity level, SIL (dB/Hz) at various microphone
distances
4.4
Sound pressure level, SPL (dB) at various microphone
distances
4.5
38
39
Time domain waveform of signal combination from two
microphones
40
4.6
Time difference between two acoustic signals
42
4.7
Sound intensity (dB) at various distances between two
microphones
4.8
Sound pressure level, SPL (dB/Hz) at various distances
between two microphones
4.9
4.10
43
44
Time domain waveform of spark sound with human
interference
46
Spark sound intensity with human interference
46
xii
4.11
4.12
Time domain waveform of spark sound with car’s sound
interference
46
Spark sound intensity with car’s sound interference
46
xiii
LIST OF TABLES
TABLE NO.
TITLE
PAGE
1.1
Gantt's chart for FYP 1
6
1.2
Gantt’s Chart for FYP 2
7
2.1
Types of shunt fault
10
2.2
Classification of cable fault with respect to Rf
3.1
and spark gap
11
List of equipments
23
xiv
LIST OF SYMBOLS & ABBREVIATIONS
Zo
-
Characteristic impedance of the cable
Rf
-
Fault resistance
Cf
-
Fault capacitance
Vp
-
Velocity of propagation
SG
-
Spark gap
TNB
-
Tenaga Nasional Berhad
MV
-
Medium Voltage
TDR
-
Time Domain Reflectometry
SIM
-
Secondary Impulse Method
AC
-
Alternating Current
DC
-
Direct Current
HV
-
High Voltage
XLPE
-
Cross-linked Polyethylene
FFT
-
Fast Fourier Transform
CHAPTER 1
INTRODUCTION
1.1
Overview of the project
Power distribution system is defined as the connection between the electric
power system and consumer. In Malaysia, there are several level voltages for
distribution system network which is 33 kV, 22 kV, 11 kV, 6.6 kV and 400/230 kV.
In distribution system, there are overhead lines and underground cables used to
deliver power to the consumer. This cable system is susceptible to the disruption just
like the other electrical system. There are many factors will cause faults in the cable
system such as aging, human error, partial discharge, down tree, and natural disaster
like lightning and landslide. This situation causes the engineers to face the
difficulties to find and detect the exact location of the cable fault especially the
underground cable.
Recently, fault location of underground cable has started to give the attention
important to the utility company in order to know the condition of their cable system,
make the quick maintenance and ensure the highest network reliability that served to
the customer. Hence, the capability to locate faults in underground cable with the
maximum accuracy and minimum time is the main concern to ensure the efficiency
of the distribution system. In order to attain the minimum time, effort and
expenditure, every cable fault should be handled with systematic procedure.
2
1.2
Problem statement
Locating a fault in underground cable systems is more difficult if compared
to the other electrical equipment. The process involves additional equipment such as
thumpers, radars, detector and many more. That equipment is used to transform the
invisibility of the cable into other forms of signals, such as acoustic sound and
electromagnetic pulses [1].
According to the book of TNB ILSAS [2] regarding cable fault location, there
are several proper procedures involved in locating faulty cables. Firstly, the
procedure is to analyze the fault occur in the cable. Secondly, the step is to
prelocation the fault using different kind of methods. The methods that used in
prelocation are bridge method, time domain reflectometry (TDR), impulse current
method or secondary impulse method (SIM). Thirdly, pinpointing the fault is to
confirm the exact position of the fault. Lastly, the confirmation of the faulty cable
and quick maintenance should be done.
However, the external interference or disturbance such as environmental
noise usually will distract procedure of locating the fault cable. It will affect and
degrade the efficiency when detecting fault on underground cables. In industrial
sites, it usually exposes to the high level of induced electrical noise. Hence it will
disrupt to the process of locating faults in MV cable. For example, a study conducted
by Easthem, Smith and Chen stated that networks with high electrically induced
noise such as process plant networks will lead the detection of partial discharge
significantly more difficult [3].
As for distribution systems, mostly route of underground cable was located
near the main road which is always noisy with the sound of the vehicles and
pedestrians across the road every day. However, whenever the fault occurs involving
underground cables, the utility should to the rapid restoration of the power supply to
the customer.
3
Unfortunately, the external interference or disturbance of the surrounding
could be distracting the procedure of locating the faulty cable especially in
pinpointing activity. This is because pinpointing should be handled manually by the
staff in charge. Hence, it could degrade the accuracy and efficiency when detecting
the fault of underground cable due to the interruption from environmental noise and
also human error.
1.3
Objectives of the project
The objectives for this project are:
1.
To investigate the effect of environmental noise in the cable fault detection.
2.
To determine the effect of pickup positions in the cable fault point.
3.
To suggest the possible solutions to eliminate the frequency of environment
noise during cable fault detection.
1.4
Scope of the project
Previous research showed the little research regarding the pinpointing method
on the electrical cables. Hence, the scope of this project is to focus more electrical
underground cable which is especially on distribution cable. Besides, the pinpointing
activity of positioning the exact location of the cable fault in a noisy environment
will be the major study in this project. The equipment of pinpointing method from
Baur product is involved to locate the fault for underground cables.
4
The shock wave discharge or acoustic method is applied in pinpointing
activity. When the surge generator repeatedly discharges to the faulty cable, an
audible sound will be produced at fault position and pickup by the detector.
In this project, the spark gap was used as representation of a cable fault
circuit. It is used to obtain the time domain waveform of the spark generated by
applying high voltage to the electrode configuration. The point-to-point electrode
configuration was employed in this project. The audible sounds of spark generated
were recorded using microphones. Therefore, the personal computer was used to
record and save directly the sound file of spark generated for analysis purpose.
Altogether the work schedules are shown in Table 1.1 and 1.2 in subchapter
1.6.
1.5
Report structure
This report consists of five chapters including this chapter. The first chapter
has some introduction about fault location of underground cable, problem statement,
objective, and scope of this project, report structure and the Gantt’s chart for both
semesters.
Chapter 2 details the theories and literature review that have been done. It
discusses on the definition of cable faults, factors and type of cable fault, procedure
of cable fault location, pinpointing activity, acoustic wave methods in faulty cable,
sound spectrum, fast Fourier transform (FFT), and stereophonic.
5
Chapter 3 discusses the methodology and instruments on completing this
project. The preparation work and experimental procedure are also presented in this
chapter. In this project, MATLAB R2010a and Praat software were used to analyze
the collected results.
Chapter 4 presents the results and discussion on the findings of this project.
The results have been presents in suitable graph and tables.
Chapter 5 describes the conclusion of this project. The suggestions for the
further works for this project are included in this chapter.
6
1.6
Gantt’s chart
Table 1.1: Gantt's chart for FYP 1
ACTIVITIES
FYP briefing
Meeting with
supervisor
Finding FYP’s title
Search related
references &
information
Confirm FYP’s title
& writing proposal
Study about the
fault cable and its
method to locate
fault
Study about
principle of
travelling wave in
cable
Study about
pinpointing method
and acoustic wave
Study about
frequency
produced by spark
in fault cable
Find equipment to
record sound of
spark
Figure out the
experimental setup
Preparation and
presentation for
seminar FYP 1
Preparation for
FYP 1 report and
submission
WEEK
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
7
Table 1.2: Gantt’s Chart for FYP 2
ACTIVITIES
Literature
review
Meeting with
supervisor
Identify the
equipments
involved
Conduct
experiment
for single
microphone
recording
Conduct
experiment
for double
microphones
recording
Analyse and
discuss the
sound of
spark
recording
Abstract
submission
for EESS15
Poster
preparation
for EESS15
EESS15
Exhibition
Report
writing
Logbook,
report draft
submission
Hard bound
report
submission
1
2
3
4
5
6
7
8
9
10
WEEK
11 12
13
14
15
16
17
18
19
20
CHAPTER 2
LITERATURE REVIEW
2.1
Introduction
Faults in cable networks are unavoidable. Increasing power consumption and
high loads, delays in capital investment, difficult environments and crowded housing
developments make it increasingly important to locate cable faults quickly and
reliably. That is a complex challenge because the combination of cause, cable type,
network structure, voltage levels and environmental conditions result in an infinite
number of possible faults. Furthermore, long outage of power supply can cause the
utility company having a heavy loss of revenue and inconvenience to the customers.
Therefore, a quick detection and identification of the faults is a major concern of the
utility company. Also, each fault will result in costs and stress. Hence, cable fault
location requires trained specialists, and they need rugged, reliable, flexible
equipment which can apply all the required methods of identifying a fault.
9
2.2
Definitions of cable fault
According to Barry Clegg in book of Underground Cable Fault Location, “A
fault can be defined as any defect, weakness, inconsistency or non-homogeneity that
affects the performance of a cable” [4]. Another definition of fault from [5] is a
physical condition that causes a component, or an element fails to perform in
the required conditions, for example, a short circuit or a broken wire. Referring to
the Oxford dictionary, fault is defined as a mistake or not perfect [6].
2.3
Causes of cable fault
Generally, cable insulation deteriorates naturally with age, especially when
exposed to elevated temperature due to high loading and even when it is not
physically damaged. Besides, there are several substances such as water, oil and
chemicals can shorten the life of insulation on the cable.
Referring to the manual by MEGGER [7], cross-linked polyethylene (XLPE)
insulation is subject to a condition termed treeing. There are several factors that
cause the presence of moisture, such as contaminants, irregular surfaces or
protrusions into the insulation plus electrical stress. These factors could be enhanced
with the inception and growth of these trees within the polyethylene material.
There are many other factors in cable failures. The frequent causes that lead
to fault are mechanical damage, mishandling cable laying, poor workmanship in
cable jointing, manufacturer’s defect, natural causes, movement of soil and
corrosion.
10
2.4
Types of cable fault
The fault that occurs on underground cable can be divided into two
categories, i.e., series and shunt fault.
A series fault is a fault involving the failure of continuity in a conductor or
cable sheath. The series fault becomes major and develops to an open circuit fault
when one or more conductors loses its continuity completely. Series fault divided
into two types of fault which are single phase open or double phase open circuit
fault.
A shunt fault is a fault that involving the failure of insulation in the
conductors. Besides, shunt fault also known as short circuit fault and it is divided
into two major categories which are balanced fault and unbalanced fault as shown in
Table 2.1. Referring to Kuan and Warwick in their journal, the most common fault
occurs in shunt type is single phase to earth fault. The shunt fault contains of void
which is occur in the insulation layer and developing to the sheath layer of the cable.
Degradation of the insulation may lead to the cable failure completely [8].
Table 2.1: Types of shunt fault
Three phase direct fault
Balanced fault
Three phases to earth fault
SHUNT FAULT
Single phase to earth fault
Unbalanced fault
Double phases to earth fault
Phase to phase fault
11
The shunt fault also can be classified into several types of fault such as low
resistance fault, high resistance fault, intermittent and flashing fault. Low resistance
faults are when the fault resistance is lower than ten times the characteristic surge
impedance of the cable. The high resistance fault has occurred when the fault
resistance is ten times bigger than cable surge impedance [8]. Intermittent fault is a
fault do not occur constantly, sometimes it is depends on the load of the cable [9].
Classification of cable fault with respect to fault resistance (Rf) and spark gap
breakdown voltage as shown in Table 2.2.
Table 2.2: Classification of cable fault with respect to Rf and spark gap
FAULT TYPE
Rf
SPARK GAP
Series
→∞
Breakdown under impulse or DC
Low resistance
< 10 Zo
Breakdown under impulse provided Rf is not too low
High resistance
> 10 Zo
Breakdown under impulse
Flashing
∞
Breakdown under impulse or DC
Intermittent
∞
Breakdown under prolonged DC
2.5
Circuit of cable fault
The cable fault can be represented by an equivalent circuit with the parallel
Rf, spark gap and Cf as shown in Figure 2.1. The values of each element can vary
widely and independently of each other. The value of Rf is depending on the degree
of carbonisation on the dielectric in the cable. Meanwhile, the value of Cf relies on
the amount of moisture present. The breakdown of spark gap is determined by the
separation of two metallic boundaries of the fault may be bridged by carbonised
insulation [10].
12
Figure 2.1: Equivalent circuit of cable fault from [2]
2.6
Procedure for cable fault location
In locating fault cable, there four steps that should be followed
systematically. The procedures are analysing the fault, pre-location, pinpointing,
confirmation and re-tests the location of the fault.
It is important to analyse the fault because all the details regarding the faulty
cable and the circumstances of the outage supply should be gathered and recorded.
The combination of insulation resistance test and continuity test should be done to
confirm the existence of the fault in the cable itself.
Prelocation of the fault is an application of a test at the terminal point of the
cable to show the indication of the distance to the fault location from the terminal
point. Nowadays, many sophisticated methods are provided to locate faults in
underground cables, ranging from bridge methods to the pulse-reflection technique
[11]. In TNB, there are four methods of prelocation that usually used such as bridge
method, time domain reflectometry (TDR) method, impulse current method, and
secondary impulse method (SIM) [3].
13
Pinpointing is an application of the test to confirm the exact location of the
fault. It is essential on direct buried cables if the location and maintenance of a
fault are to be accomplished with a single excavation. According to P.F Gale in
his book, the most of faults on high-voltage power cables are pinpointed by sensing
the acoustic signal generated when the spark gap breakdown from the application of
a voltage impulse by surge generator [10].
Confirmation and re-testing the cable fault is one step before cutting away the
faulty section from the health section of the cable. The insulation and continuity test
should be carried out again after jointing the new cable followed by a pressure test
before the supply is restored.
2.7
Pinpointing activity
Pinpointing is one of the most important steps in a cable fault location.
Besides, pinpointing is essential on the underground cables to locate and repair the
fault with a single excavation of activity. There are several reasons may lead to the
failure of pinpoint a cable fault such as confused cable route, inconsistent discharge,
inaudible acoustic sound of the fault and many more. Furthermore, the cable fault
which took several days to be located are due to the pinpointing problems [3].
Therefore, the capability to locate fault with the minimum time and higher efficiency
requires the trained and experienced staff to handle and overcome the pressure of
cable breakdowns.
Prelocation method yields the estimated distance from the starting point of
the cable to the fault location. Hence, the pinpointing step is important to determine
the exact location of the fault at the site especially for underground cable. There are
two common methods used in pinpointing which are shock wave discharge or
acoustic method and pool of potential or step voltage method.
14
The pool of potential or step voltage method is applied when cable with low
resistance fault or whenever an acoustic method cannot be implemented in this type
of fault. Using this method, ground probes is used to measure a voltage funnel which
is created in the ground [3][8].
2.8
Acoustic methods
Acoustic method or shock wave discharge is used when high voltage
capacitors from a surge generator set is repeatedly discharged into faulty cable. The
energy is dissipated at the fault and produced an audible sound and vibration within
the fault position. The Seismophone will be used to pick up the noise and vibration at
the fault position [10]. Figure 2.2 shows the acoustic sound is pickup by the acoustic
detector.
Figure 2.2: Acoustic sound produced in faulty cable from [7]
15
However, in certain cases, the acoustic sound can be detected without any
special equipment such as smell of the burnt faulty cable. Unfortunately, this method
is could consume more time and not reliable all the fault occurred. Therefore, it is
advantageous to use a ground microphone and amplifier to detect acoustic sound
produced in a faulted underground cable after injecting a pulse into it. Figure 2.3
shows the equipment of pinpointing to locate the fault position in an underground
cable.
Figure 2.3: Pinpointing equipment by BAUR company from [8]
In many years, the benefits of pinpointing cannot be denied and it is used
widely by utility company to locate faults in order to give rapid restoration of supply
to customers. But, in locating the fault using acoustic method, the prolonged
prelocation stage which is injecting pulses into the cable should be avoided because
it could be a threat to the generation of acoustic signal.
Basically, when locating faults using the acoustic method, it should handle
manually by personnel by bringing the equipment along the route of underground
cable. The approach in pinpointing is to listen carefully in the locality of fault. Then,
move the pinpointing equipment, listen to the sound and make an assumption based
16
on the heard sound. Regarding the heard sound, it can estimate whether the detector
is near or far away from the fault position.
This kind of method is approximate to the trial-and-error approach that
requires an endurance and patience to locate the fault since it involve the aspect of
time as well. In addition, when inexperienced personnel handled this equipment,
definitely, the time to locate the fault increases even more.
2.9
Audible sound
According to the Oxford dictionary, sound is defined as something that can
be heard [6]. In physics, sound is the vibration that propagates the acoustic wave of
displacement and pressure through a certain medium such as air, water or soil.
Acoustic sound usually referred as audible sound which is classified as
periodic signals of the sound observed. An audible sound having the frequency range
from 20 to 20,000 Hz where that range is capable to be heard by the average human.
The sound source could have combinations of different frequencies. Moreover, the
sound from our environment usually termed as noise where contains difference
frequencies.
Many years ago, the research on acoustic sound is extensively conducted in
many fields of engineering i.e. power engineering. Analysis of acoustic sound assists
the future developments of fault detection in high voltage equipment.
17
2.10
Sound spectrum
The sound waveform can be expressed in time domain and frequency domain
properties. Time domain is the analysis of the mathematical function with the respect
of time plot. Essentially, time domain analysis represents the real time data of a
sound signal.
However, the frequency spectrum is a preferable approach when conducting
research on audible sound. Frequency spectrum is the representation of the time
domain signal in the frequency domain. It involves the techniques of decomposing
the complex signal into the simpler signal in the frequency components.
Meanwhile, frequency analysis also adopted by researchers to improve the
technique of noise reduction in HV equipments. Referring to paper by Xuebao Li et.
al, the approach of time domain and frequency domain characteristics are used as a
tool to analyse the audible noise generated from corona source and reveal the
possibilities to eliminate the background noise appear in the waveform effectively
[15][16].
2.11
Fast Fourier Transform (FFT)
Signal processing methods widely used in the area of science and engineering
field and grow rapidly by years. Cable fault detection is an instance where the signal
processing is utilised. Signal processing methods for cable fault detection are include
Fast Fourier Transform (FFT), Short-Time Fourier Transform (STFT), and Wavelet
Transform (WT) [17].
18
Among these methods, FFT is simpler, easy to handle and effective method
to analyse different types of experimental data. Besides, Fourier transform produces
the frequency spectrum that contains the entire characteristics and information of the
original signal but in the different cast.
Practically, most of software implements the FFT in order to generate the
frequency spectrum for analysis purpose. According to book of Digital Signal
Processing by John G.P, FFT provides the mathematical approximation to the full
integral solution and also the method for computing the discrete Fourier transform of
sampled signals [18].
Referring to paper by Abdullah Kadri et.al, FFT method was implemented in
his project to detect and quantify the leakage in underground plastic pipes in water
distribution networks by collecting the acoustic signal data from the various leakage
volumes [19].
2.12
Stereophonic and binaural system
Stereophonic system is the reproducer in which two microphones feed two or
more loudspeakers to give a three-dimensional effect to the sound [20]. Stereophonic
also defined as a scheme of sound recording using two or more channels to produce
substantial effect in sound system through the two or more loudspeakers.
Binaural system is applying two microphones to the two or more independent
headphones for each observer. This system applies the ordinary both ears listening
for the original sound are directly in the listener’s ear. This system is replicated to the
normal listening [21].
19
Basically, both systems have the similarities in recording techniques. The
concept is placing the microphones for picking up single sound source for
reproduction towards a pair of loudspeaker for stereophonic while headphones for
binaural. Figure 2.4 and 2.5 below show the stereophonic and binaural system
respectively.
Figure 2.4: Stereophonic system (two pickups and two loudspeakers)
from [21]
Figure 2.5: Binaural system (two pickups and two headphones)
from [21]
Practically, the pinpointing activity in cable fault detection, it applied the
diotic system where used the one pickup device for acoustic signals and two
headphones to the listener’s ear. Figure 2.6 explains the concept of diotic system.
20
Figure 2.6: Diotic system (one pickup and two headphones)
from [21]
However, through this project, the same concept is adopted by having two
microphones to pick up the signals from the sound source. Since the objective of this
project is to observe the visual aspects of background noise within the original sound
source, therefore the usage of loudspeaker or headphones is not necessary.
2.13
Summary
All the theories and information have been explained through this chapter.
Throughout this chapter relates to the underground cable of the distribution network
and cable fault location is very essential to understand because it is very helpful to
determine the behavior of spark gap in the faulty cable. It also explains the attributes
of sound signals in time domain and frequency by implementing FFT as the signal
processing methods. Besides, at the last topic in this chapter, there is brief
information and concept regarding the recording method that employed during the
experimental works.
CHAPTER 3
METHODOLOGY
3.1
Introduction
This chapter basically shows the methodology adopted in this project. The
principal aim in this chapter is to conduct an experimental setup of measuring the
audible sound by a spark generated. During this experiment, two methods of
recording were implemented to capture the audio signal. Some software application
was used to analyse the outcomes of experiments. The flow of this project will be
discussed further in this chapter.
3.2
Project flow
The flowchart shown in this Figure 3.1 indicates the chronological order in
executing experiments in this project. Before starting the experiment, earlier study
regarding on spark generation and measurement of the audible sound on spark
discharge has been performed. And then, the equipments that used in this project are
named. The procedure and safety precaution before conducting the experimental
works has been prepared. Results obtained from the experiment should be recorded
and analyse for the discussion purpose. Lastly, the conclusion has been made to
summarize and conclude the results from this project.
22
Literature review
Equipments identification
Safety precaution and experimental setup
Generation of spark discharges
Measure audible sound of spark discharges
(Single microphone recording)
Measure audible sound of spark discharges
(Double microphone recording)
Results collection and analysis
Discussion and conclusion
Figure 3.1: Project flow
23
3.3
Equipments
Table 3.1 indicates the equipments that have been utilized throughout conducting
experiment to measure and capture the audible signal from the spark generated by
HVDC testing.
Table 3.1: List of equipments
No
Equipment
1
High voltage transformer, 0.24/100 kV, 10 kVA, 50 Hz
2
Voltage regulator
3
Voltage divider
4
Diode
5
Capacitor
6
Resistor
7
Insulator
8
Measuring spark gap
9
Point electrodes
10
Microphones
11
Digital multimeter
12
Discharge rod
13
Personal computer
All equipments that have been used in the experiment are depicted in Figure 3.2 to
Figure 3.10.
24
Figure 3.2 shows the high voltage transformer used in the experiment to
supply voltage.
Figure 3.2: High Voltage transformer
Figure 3.3 shows the voltage regulator used to increase the AC voltage until
desired value.
Figure 3.3: Voltage regulator
25
Figure 3.4 shows the diode that used as rectifier which is to convert the
output AC supply of transformer to the DC supply. The positive polarity was applied
when conducting this experiment.
Figure 3.4: Diode
Figure 3.5 shows the smoothing capacitor that used in this experiment to
supply a constant DC output.
Figure 3.5: Capacitor
26
Figure 3.6 shows the insulator used in this experiment to isolate the electrical
charged part of the equipment from another charged or uncharged metal part.
Figure 3.6: Insulator
Figure 3.7 indicate the voltage divider for HVDC testing used to measure
high voltage breakdown by displaying at digital multimeter.
Figure 3.7: Voltage divider
27
Figure 3.8 shows the measuring spark gap with point electrodes used in this
experiment to generate the spark discharge when supplying high voltage with a
constant gap distance.
Figure 3.8: Measuring spark gap and point electrodes
Figure 3.9 shows the discharge rod that used in this experiment to disperse
the static charge after de-energizing the voltage supply.
Figure 3.9: Discharge rod
28
Figure 3.10 (a) and (b) indicates the C-1U USB Studio Condenser
microphone and Probex Standing microphone respectively. These microphones used
during the experiments to measure and record the audible sound by spark discharge.
All the recording file was save in the personal computer for analysis purpose.
(a)
(b)
Figure 3.10: Microphones
For the both microphones, the frequency response is limited up to the 20 kHz.
The type of transducer used in both microphones is condenser type. While, the polar
pattern applied for the C-1U USB Studio Condenser microphone and Probex
Standing microphone are cardioid and unidirectional pattern respectively. The
detailed of the specifications of the microphones, frequency response and polar
pattern of C-1U USB Studio Condenser microphone was attached in Appendix A.
29
3.4
Generation of spark discharge
The sparks were created by supplying the high voltage in the point-to-point
electrode gap in air at atmospheric conditions. The gap distance between point
electrodes is a constant value with 2 cm. The supply voltage was increased slowly
around 26 kV to 30 kV, the spark is discharged between point electrodes. At the
same time, audible sound from the spark produced was captured by microphone and
kept into the personal computer. The point-to-point electrode configuration as
depicted in Figure 3.11.
Figure 3.11: Point electrodes configuration in spark generation
3.5
Placement of microphones
The microphone was used to measure the acoustic sound generated by the
spark. There are two different microphones that used in this experiment. The
microphone was oriented directly facing to the point-to-point electrodes. In this
project, there are two ways of recording method that had implemented through this
experiment, i.e., single microphone recording and double microphones recording.
The elaboration regarding on both methods will be discussed further in 3.5.1 and
3.5.2.
30
3.5.1
Single microphone recording
The configuration of the single microphone used was shown in Figure 3.12
and Figure 3.13. In this method, only the C-1U USB Microphone was involved. The
measurement was adopted with varying the distance from the microphone to the
sound source where the spark was generated. The distance was controlled from 0.5 m
to 3.0 m with the 0.5 m interval. For this recording, the Audacity software was used
to capture the sound waveform and saved it in the wave file.
Figure 3.12: Position of single microphone
D
Figure 3.13: Experiment of single microphone recording
31
3.5.2
Double microphones recording
The image of double microphones recording was presented as Figure 3.15.
The C-1U USB Microphone and Probex standing microphone were used to measure
the audible sound from the spark. In this experiment, the two microphones were
adopted to obtain the results. The distances between both microphones were varied
with the fixed 1.0 m distance of microphones to the audio source. Figure 3.14 show
the configuration of the measurements respectively. The Krystal Audio software was
utilized to record the spark sound from both microphones simultaneously.
Figure 3.14: Position of double microphone
R
Figure 3.15: Experiment of double microphone recording
32
3.6
Experimental setup
A high voltage transformer was used to provide AC voltage throughout the
experiment. While, the rectifier was utilized to convert the AC supply into DC
voltage. The top and bottom of the point electrodes configuration was connected to
the high voltage supply and grounded respectively. In this experiment, the electrode
gap used is fixed by the 2 cm length. Then, the microphone was setup as discussed in
previous subchapter. After setting up all the equipment, safety precautions were
adopted before and after conducting the experiment. Lastly, the voltage supply was
increased by changing the voltage regulator until the spark appeared. The
experimental work was begun with the single microphone recording followed by
double microphone recording. The spark sound has been recorded with Audacity and
Kristal software was kept to the personal computer for analysis purposes. The
analysis has been done through Maltab R2010a and Praat software. The experimental
circuit and pictorial view were depicted in Figure 3.16 and Figure 3.17 respectively.
Figure 3.16: Experimental circuit in IVAT laboratory
33
Figure 3.17: Experimental setup of double microphone recording
3.7
Wave file analysis
The experimental results were transferred and kept into the personal
computer for analysis. The spark sounds are represented in wave file to compatible
with the software used.
The Praat and Matlab R2010a software were used to analyse the sound
spectrum of each spark sound file. The FFT is a simple and effective method to
analyse experimental results. Hence, Praat software used to convert the spark sound
which is in time domain into frequency domain by using Fourier transformation. The
entire wave file for single microphone recording was analysed using this software.
For the double microphones recording, the wave file from each microphone
for each experiment should be merged into a single time domain. So, Matlab was
used to merge two different sound signals into a same timeframe. The analysis
begins by observing the time differences between both signals as the distance
between both microphones were varied.
34
3.8
Summary
In this chapter, the HVDC experiment was carried out to generate the spark
by using point-to-point electrode configuration. The primary aim in this experiment
is to measure the audible sound of the spark generated. Hence, the microphone was
used to accomplish the purpose as stated above. Two ways of recording method had
been implemented in this experiment, i.e., single microphone and double
microphones recording. During the experiment, the microphones were used to record
and save the sound file for discussion purposes. Lastly, the appropriate software was
used to analyse the experimental results.
CHAPTER 4
RESULTS AND DISCUSSION
4.1
Introduction
The results and discussion of the experimental works are presented in this
chapter. The chapter is divided into two main sections. The results from the single
microphone recording are elaborated in the first section, followed by results of
double microphones recording. Thus, the frequency domain characteristics of audible
sound from spark generated are analysed based on the time-domain results.
4.2
Single microphone recording results
4.2.1
Time-domain waveform of spark sound
Wave file of spark generated is converted into the time domain waveform by
using the Praat software. The results of the experiment are shown below by
displaying the spark waveform for each distance of the microphone to the spark
sound source. The measured audible sound by the spark generated at each distance is
presented from Figure 4.1 (a) to (f).
36
D = 0.5 m
(b) D = 1.0 m
(c) D = 1.5 m
(d) D = 2.0 m
(a)
(e) D = 2.5 m
(f) D = 3.0 m
Figure 4.1: Time-domain waveform at various distances, D
From Figure 4.1, it can be concluded that the sound pressure pulses of the
spark generated are clearly seen in the time-domain waveform. Also, it is distinctly
established that the spark waveform in time-domain contains the random time
interval between sound pressure pulses with the non-consistent bipolar amplitudes.
The background noise in the time-domain waveform showed very minimal
and not hidden the sound of the spark sound. It is can be shown that the spark sound
could be clearly picked up if there is no bigger disturbance such as car sound or
human conversation.
37
The measured time-domain waveforms above are divided into three lengths
of time. Foremost, for microphone distances of 0.5 m and 1.0 m are having
timeframe of 3 seconds, followed by timeframe of 4 seconds for microphone distance
is 1.5 m and 2.0 m and the rest is 5 seconds of time duration.
Bar chart in Figure 4.2 describes the time taken for the first spark pulse on
every distance of the microphone to the sound source. For the nearest microphone
distance, the audible sound captured by the microphone is 0.78 second while for the
others is above 1 second. In addition, bar chart showed that the time occupied by the
first pulse was increased as the distance became farther from the spark sound source.
At the 3.0 m of microphone distance indicate the decrement by 0.46 second from the
previous distance. At this distance, it showed that the pickup time is quicker than
previous measurement. This is due to the strength of the sound is louder and can
detect quickly by the microphone though the distance is far.
Acoustic sound system, the propagation of a sound wave is significantly
dispersive to surrounding especially in open space environment. This is can be
concluded that the distance from microphone to sound source in an open
environment could be influencing the travelling time of spark pulse.
Time taken for first spark pulse
3
2.76
2.84
2.38
Time (seconds)
2.5
2
1.5
1
1.14
1.18
0.78
0.5
0
0.5 m
1.0 m
1.5 m
2.0 m
2.5 m
3.0 m
Distance of microphone to sound source, D (metre)
Figure 4.2: Time taken for the first spark pulse
38
4.2.2
Intensity level of spark sound
The sound intensity level of spark sound at different microphone positions in
the time domain is shown in the Figure 4.3. From the illustrations, it can be observed
that the intensity of background noise or disturbance around the laboratory is
comparatively smaller than the spark sound intensity. Nevertheless, it also can be
noticed that the background noise is getting bigger as the microphone was placed far
from the sound source. It is can be proved when at 0.5 m, the intensity level of noise
about 16 dB. Then, the intensity approximately rose to 40 dB at 3.0 m. Thus, the
results indicated that the acoustical property of spark discharge is completely easy to
spot if the environmental noise can affect as the detector is far from the spark
location.
(a) D = 0.5 m
(c) D = 1.5 m
(e) D = 2.5 m
(b) D = 1.0 m
(d) D = 2.0 m
(f) D = 3.0 m
Figure 4.3: Sound intensity level, SIL (dB) at various microphone distances
39
4.2.3
Frequency spectrum of spark sound
The frequency spectra described for spark generated in this experiment are
depicted in Figure 4.4 (a) to (f). Those figures indicated the sound pressure level of
the spark generated in unit decibel (dB) per Hertz. It can be found that the signature
of the frequency spectrum is varied for each microphone distance to the sound
source. For the microphone distance of 0.5 m, the sound pressure level is fluctuating
and not consistent if compared to the microphone distance of 3.0 m. Therefore, it can
be concluded that the spectra for the farther microphone position was influenced by
the background noise around the laboratory that captured by microphone itself.
(a)
D = 0.5 m
(c) D = 1.5 m
(e) D = 2.5 m
(b) D = 1.0 m
(d) D = 2.0 m
(f) D = 3.0 m
Figure 4.4: Sound pressure level, SPL (dB/Hz) at various microphone
distances
40
4.3
Double microphone recording results
4.3.1
Time domain waveform of spark sound
The recording sound file from each microphone was illustrated in the time
domain waveform by combining both signals. Due to the limitation in Praat software,
the combination of two signals in one time plot cannot be performed. Thus, Matlab
was used to achieve desired works. The results of this experiment are presented
below by displaying the time domain waveform at various distances.
(a) D = 0.2 m
(c) D = 1.0 m
(e) D = 2.0 m
(b) D = 0.5 m
(d) D = 1.5 m
(f) D = 3.0 m
Figure
4.5: Time domain waveform of signal combination from two microphones
F
igure 4.5 (a) to (f) represents the combination of two signals of the same sound
41
source from various gaps between microphones. The analysis had been done by
merging the two signals into one time plot to observe the different capturing time of
both microphones towards the spark sound. The different colour was applied to
differentiate the both waveforms. The green colour represents the spark waveform
captured by C-1U USB microphone while the red colour represents spark signal from
Probex standing microphone.
The time plot for each gap between microphones was independent and not
fixed at the certain time plot. Six different gap was used begin with 0.2 m, 0.5 m, 1.0
m, 1.5 m, 2.0 m and 3.0 m. From all six figures, it showed the relationship between
both acoustic signals when capturing the spark sound. These figures depict that as the
bigger spacing between both microphones from the sound source, wider capturing
the time difference happened between two signals. So, it is indicated that the
microphones are located far from the sound source.
The sound waveform in Figure 4.5 (a) illustrated the overlapping situation
between both signals. This situation occurred due to the position of both
microphones are very near to the spark sound and located directly to the spark
location. In contrast, the waveform in Figure 4.5 (f) showed that microphones are
placed distantly from the spark location.
By observing the both captured acoustic signals, it can determine the spark
location. Besides, this phenomenon or situation can be conformed to the real practice
in detecting the acoustic signal in spark discharge by the energized faulty cables. It is
also presented by Ju-Chu Hsieh et al. in their paper, it is shown that the difference
between two acoustic waves would indicate the nearness of the detector to the defect
location [22].
42
Time, ∆t (seconds)
Time difference between two signals
0.2
0.166
0.15
0.1
0.057
0.05
0.001
0
0.065
0.023
0
0.2 m
time (seconds)
0.5 m
0.2 m
0
0.5 m
0.001
1.0 m
1.0 m
0.023
1.5 m
1.5 m
0.057
2.0 m
2.0 m
0.065
3.0 m
3.0 m
0.166
Distance between microphones, D (metre)
Figure 4.6: Time difference between two acoustic signals
Bar chart in Figure 4.6 shows the time difference between two acoustic
signals at a certain distance between two microphones. The data are tabulated
systematically from the closest distance to the widest distance between two
microphones.
It can be clearly noticed that there is no time difference for both acoustic
signals for the 0.2 m. Refer to Figure 4.5 (a), it is indicated that the green signal is
completely on the red signal and there is no time lagging when capturing signals
from the audible spark sound. At distance 0.5 m, red signal is delayed by 0.001
second with green signals. However, both signals do not face too much difference in
terms of time domain waveforms. For distance of 1.0 m, 1.5 m, 2.0 m and 3.0 m, it is
showing the significant value in seconds when picked up signals from both
microphones simultaneously. The delaying time is increasing as the microphones gap
bigger as shown in the figure above. Moreover, the 3.0 m gap produced 0.166 second
in delaying time for both signals and it is hugely different from the preceding
microphone gaps.
Hence, time difference increases as the farthest distance between
microphones. It is can clearly observe through data and waveforms in Figure 4.5 and
Figure 4.6 respectively.
43
4.3.2
Intensity level of spark sound
Figure 4.7 presents the sound intensity level of spark sound with different
microphones gaps. Those figures show that the level background noise captured is
quite alike for each distance approximately at 60 dB level. It shows that background
noise around laboratory will not gave significant impact when two microphones were
utilised to pick up spark sound. The contrast from single microphone results is the
distance of microphone to spark source is not affecting the intensity level of
background noise. Even though the decibel level is high, practically the spark sound
can be heard in normal situation without any noisy interruption.
(a) D = 0.2 m
(b) D = 0.5 m
(c) D = 1.0 m
(d) D = 1.5 m
(e) D = 2.0 m
(f) D = 3.0 m
Figure 4.7: Sound intensity (dB) at various distances between two
microphones
44
4.3.3
Frequency spectrum of spark sound
Figure 4.8 (a) to (f) represents the frequency spectrum of the spark generated
for double microphone recording experiment with the various gap distances between
microphones. It can be noted that the pattern of the frequency spectra is similar for
each distance but different in terms of pressure levels in dB. Besides, the sound
pressure level of spark is becoming weaker as the gap distance of microphones
farther. In normal situation, these outcomes also can show the identification of spark
location by referring to the pressure levels.
(a) D = 0.2 m
(b) D = 0.5 m
(c) D = 1.0 m
(d) D = 1.5 m
(e) D = 2.0 m
(f) D = 3.0 m
Figure 4.8: Sound pressure level, SPL (dB/Hz) at various
distances between two microphones
45
4.4
Spark sound with interference
The concern of this project is to observe and investigate the existence of
interference when detecting faulty whereas audible spark sound is generated. The
noise may distract the detection work due to acoustical characteristic by passing
vehicle sound, human conversation, natural sound and environmental noise by
factories. These possible factors could be hiding the actual acoustic sound of spark
when detecting fault. Hence, to verify the effect of interference, the analysis had
been done when disturbance from human and passing car are appearing in the spark
sound. The analysis is performed in time domain only.
4.4.1
Spark sound with human interference
Figure 4.9 and 4.10 shows the spark waveform of amplitude and intensity
properties in the time domain plot respectively. The waveform above was influenced
by the human interference during the experimental work was conducted. Microphone
distance of 3.0 m was chosen in this situation to observe the effect of noise and only
one microphone was used to record the audible sound occurred during the
experiment.
The first pulse in the waveform in Figure 4.9 indicates the pulse that obtained
from the footstep of a person going into the laboratory area. From the amplitude
properties, it is observed that the peak of interruption pulse is smaller than spark
pulse and still showed the distinction between it. Still, the intensity level of
interference and spark pulse are quite high and similar approximate at 67 dB. It could
be an interruption when identifying and sorting whether it is spark pulse or not.
46
Figure 4.9: Time domain waveform of
spark sound with human interference
4.4.2
Figure 4.10: Spark sound intensity
with human interference
Spark sound with car’s sound interference
Figure 4.11 and 4.12 illustrates the waveform from the spark discharge and
passing car sound in amplitude and intensity properties respectively. The both
waveform was recorded separately but at the same time plot with a certain distance
of microphone position as depicted in Figure 4.11. The main purpose of this result is
to show the implication of the interference sound towards the spark waveform in
terms of amplitude and intensity level.
Figure 4.12 shows the combined intensity level of both waveform and it is
clearly seen that the spark intensity level is completely absorbed and immersed into
the car sound. It is can be concluded that when the passing car waveform is merging
into spark waveform, it could be as a disturbance to spark discharge.
Figure 4.11: Time domain waveform of
spark sound with car’s sound interference
Figure 4.12: Spark sound intensity with
car’s sound interference
47
4.5
Summary
The results presented in this chapter support the objectives in this project to
observe the effect of environmental noise to the faulty cable where represented it
with spark gap. From the results, it can be concluded that spark waveform could be
influenced by background noise as the pickup device which is microphone placed
farther from the spark source. Besides, the double microphone recording was
conducted to observe the delaying time captured by both microphones. Also, the
overlapping signals in time domain waveform could indicate the spark location came
from and as well as shows the noise cancellation between both microphones. Next
chapter will discuss the conclusion and recommendation for this project.
CHAPTER 5
CONCLUSION AND RECOMMENDATIONS
5.1
Conclusion
The acoustic detection of faulty underground cable is important to locate the
fault point along the cable route. The underground cable route usually located nearby
to the main and noisy roads. Therefore, the sound of passing vehicles could be
interrupting the acoustic detection in faulty cable. Hence, the characteristic of cable
fault was represented by an equivalent circuit of spark gap which was conducted in
this experiment.
The results present the time domain and frequency domain properties of the
spark generated by using one microphone and two microphones to capture the
audible sound by the spark. The various distances of microphone were set up to
observe the existence of background noise in the laboratory towards the captured
signals.
49
For single microphone recording results, it is can be noticed that when the
position of microphone is far from the spark point, the time taken for audible sound
to be captured by the microphone is increased. Still, the intensity level of background
noise also showed the increment pattern as the distance is increased.
For double microphones recording results showed that the both captured
signals were superimposed when the closest microphone gap is applied which is 0.2
m. However, the time difference between signals was larger as the distance
increased. The intensity level of background noise shows the different outcomes
compared to the single microphone recording. From the Figure 4.7 (a) to (f), the
intensity level of background noise shows the level at 60 dB at each distance.
Although, that level is higher, still, there is no significant increment for every
distance.
The analysis regarding spark sound with human and passing car interference
show the significant results of environmental noise in spark sound. The intensity of
footstep of human could be higher as much as the loudness level of spark pulse. It
means the human interruption could be distracting the identification of spark pulse if
the spark sound and disturbance occur at the same time. It is can be demonstrated by
the results of passing car interference into the spark sound. According to the intensity
level of both signals, the intensity of interference and spark pulse is completely
absorbed and the spark pulse cannot identified then.
By conducting this experiment, the intensity of background noise could be
compared with the increment of sound intensity level of noise in each microphone
position as proved in single microphone recording. Besides, the effect environmental
noise was convinced by the sound of passing car and human interruption into the
recorded spark sound. Therefore, it can be concluded that the background noise from
the environment could be effect in configuring the spark sound in the cable fault
detection.
On the other hand, the placement of the pickup device could be influenced
the results of measuring audible sound of spark. It can be showed in both
experiments whereas the distance of microphones caused the travelling time of sound
50
wave propagation become bigger as the distance wider. The results of two
microphones recording showed that the lagging time for both microphones to capture
sound was increased sequentially with microphone position. Therefore, the position
of pickup device showed the consequence in determining the cable fault point.
Lastly, the outcome of this experiment was attained to achieve the objectives
of this project. The evidence where the environmental noise and position of pickup
device were proven could give effect in detecting acoustic signal of faulty cable.
Hence, it is anticipated that results the results presented in this report can be worthy
and beneficial information regarding to the spark sound properties to the user. Hence,
the improvement in identification of the faulty cable should be done in order to have
a reliable power system network.
51
5.2
Recommendations
In this project, the timeline of the captured spark sound is random and nonuniform. Thus, it is highly suggested to conduct the experiment within the constant
timeline to obtain the reliable results of the experiment. Besides, in order to reduce
the error of captured signals for double microphone recording, it is recommended to
use the same type of microphones. By using the same kind of microphone, the error
of capturing signal can minimize since the same properties and specifications of
microphones are used.
For the analysis process, there is one problem encountered while attempting
to do conversion of extension wave file into csv file which is can be read and
analysed by MS-Excel. Therefore, it is recommended to use Java or C programming
to solve the problem above. By performing the task, the analysis by plotting two
different signals into a timeline could easier. Besides, the advanced signal processing
may be used such as wavelet transform to observe the comparison with the FFT
results.
52
REFERENCES
[1]
M.T Glinkowski; N.C. Wang. “ANNs pinpoint underground distribution
faults”. IEEE Computer and Applications in Power. October 1995.
[2]
TNB ILSAS-Training division, “Cable fault location”.
[3]
Carl Easthem; Dr Colin Smith; Dr Fa-Chung Chen. “Detection and location
of PD in MV cables in electrically noisy industrial environments”. 21st
International Conference on Electrical Distribution. June 6-9, 2011.
Frankfurt : CIRED, 2011.
[4]
Barry Clegg, “Underground cable fault location”, Mc-Graw Hill, 1993.
[5]
T.A Short, “Electric power distribution handbook”, CRC PRESS, 2004, pp
311-312.
[6]
Miranda steel, “New Oxford Dictionary”, Oxford Fajar sdn Bhd, 2009, pp
305.
[7]
MEGGER company, “Fault finding solutions” application notes.
[8]
K.K. Kuan; Prof. K. Warwick, “Real-time expert system for fault location on
high voltage underground distribution cables”, IEEE Proceedings-C, Vol.
139, No. 3, May 1992.
[9]
BAUR company, “BAUR cable fault location” application notes.
[10]
M. A. Laughton; M.G Say, “Electrical Engineer’s Reference Book”, 14th ed.
Butterworth International Editions, 1985.
53
[11]
B.Clegg and N.G.Lord, “Modern cable-fault-location methods”, IEEE
Proceedings, April 1975, Vol 122(4).
[12]
Jim Livie; Philip Galet; Anding Wang, “The Application of on-line travelling
wave techniques in the location of intermittent faults on low voltage
underground cables”, pp 714-719.
[13]
Ajendra Dwivedi; Xinghuo Yu, “Fault Location in Radial Distribution Lines
Using Travelling Waves and Network Theory,” Power Engineering Society,
2011, pp 1051-1056.
[14]
C. Ayrault, P. B´equin and S. Baudin, “Characteristics of a spark discharge as
an adjustable acoustic source for scale model measurements”, Proceedings of
the Acoustics 2012 Nantes Conference, April 23-27, 2012. Nantes, France :
2011, pp 3549-3553
[15]
Xuebao Li , Xiang Cui , Tiebing Lu , Di Zhang and Yang Liu, “ Timedomain Characteristics of the Audible Noise Generated by Single Corona
Source under Positive Voltage”, IEEE Transactions on Dielectrics and
Electrical Insulation, April 2015, Vol. 22(2).
[16]
Xuebao Li, Xiang Cui, Tiebing Lu, Di Zhang, Zhenguo Wang and Wenzuo
Ma, “Time-domain and Frequency-domain Characteristics of Audible noise
from Single Corona Source”, International Conference on Power System
Technology, October 20-22, 2014. Chengdu : Powercon, 2014,
[17]
Abhishek Pandey; Nicolas H. Younan, “Underground Cable Fault Detection
and Identification via Fourier Analysis”, IEEE Journal Paper, 2010, pp 618621.
[18]
John G. Proakis; Dimitris G. Manolakis, “Digital Signal Processing
Principles, Algorithms and Applications”, 3rd ed, Prentice Hall International,
1996.
54
[19]
Abdullah
Kadri;
Elias
Yaacoub;Mohammed
Mushtaha,
“Empirical
Evaluation of Acoustical Signals for Leakage Detection in Underground
Plastic Pipes”, 17th IEEE Mediterranean Electrotechnical Conference, April
13-16, 2014. Beirut, Lebanon : IEEE, 2014.
[20]
Jin-Young Park; Ji-Ho Chang; Yang-Hann Kim; Youngjin Park, “Personal
Stereophonic System Using Loudspeakers: Feasibility Study”, International
Conference on Control, Automation and Systems, October. 14-17, 2008.
Seoul, Korea : COEX, 2008.
[21]
William B. Snow, “Basic Principles of Stereophonic Sound”, Journal of
SMPTE IRE Transactions – Audio, vol. 6.
[22]
Ju-Chu Hsieh; Cheng-Chi Tail; Ching-Chau Su; Chien-Yi Chen; Jiann-Fuh
Chen; Yu-Hsun Lin, “The Application of Partial Discharge Detector and
Electro-Acoustic Signals Analysis Methods for Power Cables Monitoring”,
IEEE International Conference on Condition Monitoring and Diagnosis,
September 23-27, 2012. Bali, Indonesia : IEEE, 2012, 157-160.
[23]
G. C. Sibilant, A. C. Britten, W De Villiers, “Determination of the Noise
Signatures of Various Sparkover Phenomena on the Insulated Shield Wire of
Long Hvdc Lines”, Inaugural IEEE PES Conference and Exposition, July 1115, 2005. Durban, South Africa : IEEE, 2005, pp 423-427.
55
APPENDIX A
SPECIFICATIONS OF MICROPHONES
C-1U USB Studio Condenser Microphone
Probex Standing Microphone
56
FREQUENCY RESPONSE AND POLAR PATTERN OF C1-U USB
MICROPHONES
Frequency Response of C-1U USB Studio Condenser Microphone
Cardioid Polar Pattern of C-1U USB Studio Condenser Microphone
57
EXPERIMENTAL VIEW ON DOUBLE MICROPHONES RECORDING
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF

advertising