1

2
 DEGRADATION OF POLYMERIC POWER CABLE DUE TO WATER TREE
UNDER DC VOLTAGE

MUHAMAD IHSAN BIN SUJANI

A project report submitted in partial fulfillment of the requirements for the award of
Bachelor of Engineering (Electrical)

Faculty of Electrical Engineering

Universiti Teknologi Malaysia

APRIL 2010
ii
 ii

Dedicated in thankful appreciation for support, encouragement and understandings

To:

My supervisor Dr.Yanuar Zulardiansyah Arief;

My beloved dad Sujani bin Samingon and my caring mom Sarifah bte Dol ;

My brothers and sisters;

Also my friends and whoever helps me to contribute into this project

 iii

ACKNOWLEDGEMENT

First and foremost, I would like to express my heartily gratitude to my

supervisor, Dr.Yanuar Zulardiansyah Arief for the guidance and enthusiasm given
throughout the progress of this project.

My appreciation also goes to my family who has been so tolerant and supports
me all these years. Thanks for their encouragement, love and emotional supports that
they had given to me.

I would also like to thank our Control 1 Lab Assistant, En. Anuar Bin
Kamarudin, En. Zamri Bin Kassim, and En. Hairoisyam Bin Abd Rani for their
cooperations, guidance and helps in this project.

Nevertheless, my great appreciation dedicated to my beloved friends and class

member’s and those whom involve directly or indirectly with this project. There is no
such meaningful word than…..Thank You So Much.
 iv

ABSTRACT

The majorities of XLPE insulated cables installed in the ground are exposed to
moisture and as a result are susceptible to degradation due to water trees. Water trees
are generally found to initiate and grow in XLPE insulation exposed to an alternating
electric field and humidity. Impurities inside the insulation material will increase the
risk of water tree initiation.

This project focused on study of degradation of polymeric power cables due to

water trees under High Voltage Direct Current ( HVDC ). Three of polymeric cables
samples which non-defect condition, medium defect condition and very defect condition
will be immersed in the container filled by water. Then the cables samples were injected
with 2 kV HVDC. The electrical parameters such as tan δ, power factor, insulation
resistance, insulation capacitance and voltage recovery profile were measured and
analysis to study a correlation between the water trees and insulation degradation on
XLPE insulation.

As a result it shows the value of insulation resistance is lower for severe defect
condition than moderate defect condition. Tan δ and power factor shows severe defect
condition have higher reading value than moderate defect condition. For capacitance,
the increasing of defect level on insulation give higher reading value of capacitance.
Recovery voltage profile shows a correlation between ageing time and ageing level.
 v

ABSTRAK

Majoriti kabel berpenebat XLPE yang digunakan dibawah tanah terdedah kepada
kelembapan dan kualiti penebat boleh berkurang disebabkan oleh water trees. Water
trees biasanya ditemui didalam kabel berpenebat XLPE yang terdedah kepada medan
elektrik dan kelembapan. Bendasing yang terdapat didalam penebat juga boleh
meningkatkan risiko untuk water trees terbentuk.

Projek ini memfocuskan unruk mengkaji pengurangan kualiti penebat kabel

disebabkan oleh water trees pada medan elektrik arus terus. Tiga sampel kabel
digunakan dalam projek ini, setiap kabel mempunyai keadaan penebat yang berbeza
iaitu penebat yang tidak rosak, penebat dengan sedikit rosak dan penebat yang rosak
teruk. Sampel ini kemudiannya direndam didalam air dan dikenakan medan elektrik
arus terus. Parameter elektrik seperti tan δ, faktor kuasa, rintangan penebat, capacitor
penebat dan voltage recovery profile diukur serta analisis dilakukan untuk mengkaji
hubungan antara pembentukan water trees dan pengurangan kualiti penebat pada XLPE.

Daripada kajian ini menunjukkan rintangan penebat lebih rendah untuk sampel
kabel yang rosak teruk berbanding sampel kabel yang rosak sedikit. Tan δ dan power
factor pula menunjukkan sampel kabel yang rosak teruk member bacaan yang lebih
tinggi berbanding sampel kabel yang rosak sedikit. Bagi nilai capacitor, peningkatan
tahap kerosakan pada sampel kabel memberikan bacaan yang lebih tinggi. Bacaan
voltage recovery profile pula menunjukkan hubungan diantara masa ageing dan tahap
ageing.
 vi

TABLE OF CONTENTS

CHAPTER TITLE PAGE

DECLARATION OF THESIS ii

DEDICATION iii

ACKNOWLEDGEMENT iv

ABSTRACT v

ABSTRAK vi

TABLE OF CONTENT vii

LIST OF TABLES x

LIST OF FIGURES xi

LIST OF ABBREVIATION xiii

LIST OF APPENDICES xv

1 INTRODUCTION

1.1 Background 1
 vii

1.2 Problem Statement 2

1.2 Objective of Project 2

1.4 Scope of Project 2

2 THEORY AND LITERATURE REVIEW

2.1 Water Treeing 4

2.2 High Voltage Direct Current 9

2.3 Cross Linked Polyethylene (XLPE) 12

3 METHODOLOGY

3.1 Introduction 14

3.2 Experiment Procedures 15

3.3 Measurement Setup 17

3.3.1 Measurement of Insulation Resistance 17

3.3.2 Measurement of Capacitance, Dissipation 19

Factor (tan δ), and Power Factor

3.3.3 Measurement of Recovery Voltage Profile 19

4 CHAPTER 4

4.1 Introduction 21

4.2 Insulation Resistance 22

 viii

4.3 Tan δ and Power Factor 24

4.4 Capacitance 26

4.5 Recovery Voltage Profile 28

5 CHAPTER 5

5.1 Conclusion 33

5.2 Recommendation 34

REFERENCES 35

APPENDICES 37
 ix

LIST OF TABLES

TABLE TITLE PAGE

Table 1 Resistances (GΩ ) 37

Table 2 Tan δ 38

Table 3 Power Factor 39

Table 4 Capacitance (pF) 40

Table 5 Cable 1 ( Severe Defect Condition ) 20 Hours 41

Table 6 Cable 2 ( Medium Defect Condition ) 20 Hours 42

Table 7 Cable 3 ( Non Defect Condition ) 20 Hours 43

Table 8 Cable 1 ( Severe Defect Condition ) 40 Hours 44

Table 9: Cable 2 ( Medium Defect Condition ) 40 Hours 45

Table 10 Cable 3 ( Non Defect Condition ) 40 Hours 46

 x

LIST OF FIGURES

FIGURES TITLE PAGE

Figure 2.1 Vented Water Trees 6

Figure 2.2 Bow-Tie Water Trees 6

Figure 2.3 Cable wafer with extensive voids 7

Figure 2.4 Variation of tree density versus aging time 8

Figure 3.1 Power Cable Immersed in water 15

Figure 3.2 Constructed Experiment Setup 16

Figure 3.3 Megger Meter 18

Figure 3.4 XLPE Cables Sample 18

Figure 4.1 Resistance vs Time 22

Figure 4.2 Tan δ vs Time (Hours) 24

Figure 4.3 Power Factor vs Time (Hours) 25

Figure 4.4 Dissipation Factor Model 26

Figure 4.5 Capacitance(pF) vs Time(Hours) 27

Figure 4.6 Recovery Voltage (Non Defect) 29

 xi

Figure 4.7 Recovery Voltage (Moderate Defect) 30

Figure 4.8 Recovery Voltage (Severe Defect) 30

Figure 4.9 Recovery Voltage (20 Hours) 31

Figure 4.10 Recovery Voltage (40 Hours) 31

Figure 5 Tettex Instruments – Schering Bridge Model 2816 47

Figure 6 Sample connection to Tettex Instruments – Schering 48

Bridge Model 2816

Figure 7 Tettex instrument type 5461Recovery 49

Voltage Meter
 xii

LIST OF ABBREVIATION

AC - Alternating Current

DC - Direct Current

cm - Centimeter

E - Ground

EBA - Ethylene-Butyl-Acrylate

EEA - Ethylene-Ethyl-Acrylate

EHV - Extra High Voltage

EVA - Ethylene- Vinyl- Acetate

GΩ - GigaOhm

HV - High Voltage

HVDC - High Voltage Direct Current

Ic - capacitor current

Ir - Leakage current

kV - KiloVolt

LDPE - Low Density Polyethylene

 xiii

LLDPE - linear low density polyethylene

MDPE - Medium Density Polyethylene

PE - Polyethylene

pF - pikoFarad

R - Insulation resistance

RMS - Root Mean Square

RVM - Recovery Voltage Meter

tan δ - dissipation factor

Tc(s) - time charging in second

Td (s) - time discharge in second

U - test voltage

UrMAX - peak test voltage

V - Injected voltage

XLPE - Cross Link Polyethylene

 xiv

LIST OF APPENDICES

APPENDIX TITLE PAGE

Appendix A Electric Parameters 37

Appendix B Recovery Voltage Profile Result 41

Appendix C Measurement Setup 47

 1

CHAPTER 1

PROJECT OVERVIEW

1.1 Background of study

In the early 1970s, the first reports of increasing failures on PE-insulated,

medium-voltage cables were noted in the United States. At the beginning of the 1980s,
similar cable faults were noted also in Europe. The cause of these faults was found to be
the formation of ‘water trees.’ It was found that water trees grow with the correct
combination of water, imperfections of the PE, XLPE or EPR insulation around a cable
and the operating voltage over the course of time [1].

When a water tree reaches a critical size, it can convert into an electrical tree
accompanied by partial discharge to ground that inevitably leads to an electrical fault on
the cable. All solid dielectric medium-voltage cables are susceptible to water treeing
regardless of the existence of a jacket or strand filling [1].
 2

1.2 Problem Statement

More underground power cable will be used in power system. This is due to a
request about urban landscape and site land restrictions. Underground power cable will
be exposed to a soil moisture and humidity. This condition will lead water treeing in
underground power cables. Commonly most of premature breakdown of underground
power cable is cause by water treeing.

1.3 Objective

The objectives of this project are:

1. To study the degradation of polymeric power cable due to water tree under
HVDC.

2. To determine electrical parameters such as tan δ, insulation resistance, insulation

capacitance, and voltage recovery on polymeric power cable due to water tree
under HVDC

1.4 Project Scope

This project covers the electrical characteristic such as tan δ, insulation

resistance, insulation capacitance, and voltage recovery on polymeric power cable.
 3

Power cables cross linked polyethylene (XLPE) rated 1 kV will be used as specimen.
The specimen will be injected with 2 kV high voltage direct current for 40 hours.
 4

CHAPTER 2

LITERATURE REVIEW

2.1 Water Treeing

Water treeing is a well known phenomena and a lot of papers have been
published, devoted principally to the phenomenon but there is not yet a comprehensive
model satisfying the networks engineers interested in the remaining lifetime of their
cables characterized by their own morphology due to manufacturing. Many parameters
affect the growth phase: the field frequency and value, the temperature, the polymer
morphology (chemical composition, crystalline and amorphous phases), and the ions
nature [2].
 5

It is well known that the electrical properties of insulating materials used with
high voltage applications are altered in time. This phenomenon, called "ageing", this
strongly related to the external factors acting upon the materials (electric field,
temperature, etc.) and to the manufacturing process. Over the last decades, considerable
efforts have been made to understand and to prevent ageing [2].

An insulating material is supposed electrically neutral; however, electric charges

can penetrate within the material. It has been shown that the accumulation of electric
charges (space charge) within insulation, by creating a supplementary internal field
(remnant electric field), is one of the main causes of ageing. Indeed, it seems that the
more insulation stores space charge, the more its ageing is accelerated by increase of the
global electric field (sum of the applied field and the remnant field). These charges can
cause significant field distortion to affect the insulator performance and to reduce its
lifetime [2].

Water trees are small discrete voids separates by insulator. Water tress develops
slowly, growly over a period of month or years. Much less electrical stress is needed to
cause water trees. Water trees actually look like fans, bloom or bushes.

There are essentially two types of water trees, vented water trees [Figure 2.1] and
bow-ties water trees [Figure 2.2]. Vented water trees that originate from surface of the
insulation and are potentially the most dangerous

Vented trees are trees which initiate at an electrode insulation interface and grow
towards the opposite electrode. Having access to free air is a very important factor for
the growth of the vented trees. These trees are able to grow continuously until they are
long enough to bridge the electrodes, therefore causing failure in the insulation [3].
 6

Figure 2.1: Vented Water Trees [4]

Figure 2.2: Bow-Tie Water Trees [5]

 7

Figure 2.3: Cable wafer with extensive voids [6]

The bow-tie water trees grow from contamination or voids [Figure 2.3] within
the insulation and tend to grow to a limited size without breaching the insulation. Bow-
tie trees are trees which start to grow from within the dielectric insulation and grow
symmetrically outwards from the electrodes. As the trees start within the insulation,
they have no free supply of air which will enable continuous support of partial
discharges [3]. The branches of water trees actually appear to comprise a high density
of water filled voids of typical diameter 1-10µm [7].

Research showed that water trees increased in size over time. M.I.Qureshi –
conduct experiment of resistance XLPE insulation toward water tree growth. The
experiment revealed that water tree density varies with ageing time that follow a log-
linear relation as shown in Figure 2.4.
 8

Figure 2.4: Variation of tree density versus aging time [8].

Ideally, polyethylene insulation is hydrophobic and homogeneous. In reality,

though, a large percentage of a cable is insulation consists of amorphous, non-crystalline
material with voids and interfaces that have a polar hydrophilic nature. Water and ionic
impurities are attracted to these areas and readily diffuse into the insulation,
concentrating in these polar areas. Once there, under the influence of the cable is
electrical field, water can be disassociated into damaging radicals[9].

Equations (1) and (2) are examples[9].

OH- (from water) → OH (hydroxyl group) + e- (1)

 9

2H2O → H2O2 (hydrogen peroxide) +2H+ + 2e- (2)

These radicals are damaging to the polyethylene structure. They oxidize

(degrade) the interfacing surfaces, further polarizing them and thereby propagating water
tree growth. This explanation of water tree propagation requires the presence of three
factors: amorphous polar imperfections in the dielectric structure; a sufficiently strong
alternating-current electrical field and, of course, water [9].

The formation of water trees does not necessarily mean the cables will fail. A
water trees can even bridge the entire dielectric without immediate failure occurs.
Failure occurs when water trees convert to electrical tress [10].

2.2 High Voltage Direct Current

A high-voltage, direct current (HVDC) electric power transmission system uses

direct current for the bulk transmission of electrical power, in contrast with the more
common alternating current systems. For long-distance distribution, HVDC systems are
less expensive and suffer lower electrical losses. For shorter distances, the higher cost
of DC conversion equipment compared to an AC system may be warranted where other
benefits of direct current links are useful.

The advantage of HVDC is the ability to transmit large amounts of power over
long distances with lower capital costs and with lower losses than AC. Depending on
voltage level and construction details, losses are quoted as about 3% per 1,000 km.
 10

High-voltage direct current transmission allows efficient use of energy sources remote
from load centres[11].

In a number of applications HVDC is more effective than AC transmission.

 Undersea cables, where high capacitance causes additional AC losses.

 Endpoint-to-endpoint long-haul bulk power transmission without intermediate
'taps', for example, in remote areas
 Increasing the capacity of an existing power grid in situations where additional
wires are difficult or expensive to install
 Power transmission and stabilization between unsynchronised AC distribution
systems
 Connecting a remote generating plant to the distribution grid.
 Stabilizing a predominantly AC power-grid, without increasing prospective short
circuit current
 Reducing line cost. HVDC needs fewer conductors as there is no need to support
multiple phases. Also, thinner conductors can be used since HVDC does not
suffer from the skin effect
 Facilitate power transmission between different countries that use AC at
differing voltages and/or frequencies
 Synchronize AC produced by renewable energy sources

Long undersea cables have a high capacitance, since the conductors are
surrounded by a relatively thin layer of insulation and a metal sheath. The geometry is
that of a long co-axial capacitor. Where alternating current is used for cable
transmission, this capacitance appears in parallel with load. Additional current must
flow in the cable to charge and discharge the cable capacitance, which generates reactive
power and additional resistive losses in the conductors of the cable.
 11

Additionally, there is a dielectric loss component in the material of the cable

insulation, which consumes power. When direct current is used, the cable capacitance is
only charged when the cable is first energized or when the voltage is changed; there is
no steady-state additional current required to charge and discharge the cable capacitance.
For a long enough undersea cable, the entire current-carrying capacity of the conductor
could be used to supply the charging current alone. This limits the length of AC cables.
DC cables have no such limitation. However, some DC leakage current continues to
flow through the dielectric; this is very small compared to the cable rating.

HVDC can carry more power per conductor because, for a given power rating,
the constant voltage in a DC line is lower than the peak voltage in an AC line. In AC
power, the root mean square (RMS) voltage measurement is considered the standard, but
RMS is only about 71% of the peak voltage. The peak voltage of AC determines the
actual insulation thickness and conductor spacing. Because DC operates at a constant
maximum voltage, this allows existing transmission line corridors with equally sized
conductors and insulation to carry 100% more power into an area of high power
consumption than AC, which can lower costs [11].

Because HVDC allows power transmission between unsynchronized AC

distribution systems, it can help increase system stability, by preventing cascading
failures from propagating from one part of a wider power transmission grid to another.
Changes in load that would cause portions of an AC network to become unsynchronized
and separate would not similarly affect a DC link, and the power flow through the DC
link would tend to stabilize the AC network. The magnitude and direction of power
flow through a DC link can be directly commanded, and changed as needed to support
the AC networks at either end of the DC link. This has caused many power system
operators to contemplate wider use of HVDC technology for its stability benefits alone.
 12

2.3 Cross Linked Polyethylene (XLPE)

Polyethylene is long chain hydrocarbon molecules manufactured by the

polymerization of ethylene gas. The main chemical elements making up polyethylene is
carbon and hydrogen. The elements are combined by covalent chemical bond and the
polymer structure has low polarity which gives inherent good electrical characteristics
as low electrical losses.

Polyethylene is manufactured from petrochemicals as oil/nafta or gas and

recently there have also been studies to use biomass as raw material. Polyethylene is a
very versatile polymer with a lot of different end applications. Polyethylene is
commonly classified based on density and melt flow rate which gives some of the
basic characteristics. In this specification type refers to the density range and
category refers to melt flow range. There are different polymerization processes for
manufacturing of polyethylene and these different processes give somewhat
different characteristics to the polymer.

The high pressure process will manufacture low density polyethylene,

LDPE and copolymers as ethylene-ethyl-acrylate, EEA, ethylene-butyl-acrylate,
EBA and ethylene- vinyl- acetate, EVA, respectively. The LDPE is the preferred
material for XLPE insulation and copolymers as EBA and EEA are used as base resins
for bonded conductive screens. EVA may be used as base resin for bonded as well as
strippable insulation screens [12].

Low pressure processes uses catalyst and these processes are suitable for
manufacturing of linear materials as linear low density polyethylene, LLDPE,
medium density polyethylene, MDPE and high density polyethylene, HDPE. The
robustness of particularly MDPE and HDPE makes these types of materials
suitable as jacketing materials of cables.

The crosslinking is a chemical reaction where polymer chains are linked

together forming a three dimensional network. The link formed between the
 13

polyethylene molecules is a covalent chemical bond ant it is practically an irreversible

reaction i.e. the polymer cannot be molten and formed to a different shape. Polymers
containing a three dimensional network is sometimes referred to as thermo-sets, while
polymers that can be molten and reshaped are referred to as thermoplastic materials [12].

Crosslinking of low density polyethylene can be performed by different technologies as

· Peroxide cross linking

· Silane cross linking

· Irradiation cross linking

Peroxide crosslinking is by far most the common practice for medium

voltage (MV) and the only technology applied in high voltage (HV) and extra high
voltages (EHV) cable manufacturing. Silane cross linking is the dominating
technology for manufacturing of low voltage XLPE cable. Irradiation crosslinking are
mainly used in special applications where the thickness of the polymer layer is relatively
thin[12].
 14

CHAPTER 3

METHODOLOGY

3.1 Introduction

This chapter will discuss a methodology for conducting this project. It divides
into two parts, experiment procedures and measurement setup. Experiment procedures
consist of preparing samples, constructing experiment setup and experiment process.
Methodology of measurement procedures for this study is focus on measurement of
electrical parameters such as insulation resistance, tan δ, power factor, capacitance and
voltage recovery profile of aged XLPE cable insulations due to water tree under DC
voltage. The measurement were obtain by using Megger ohm meter for measuring
insulation resistance parameter while Tettex Instruments – Schering Bridge Model 2816
has been used for tan δ, power factor and capacitance parameters measurement.
Furthermore, the Tettex type 5461 Recovery voltage meter has been used to measure the
recovery voltage profile parameter of XLPE underground cable.
 15

3.2 Experiment Procedures

1. Both end of cables jacket is removed by using metal saw and knife.

2. Then a nails is placed in the core of the both end cables. This is for connection
between DC supply and cables.

3. Metal shield on the both cables is not all been removed. This will be used for
grounding cables.

4. For creating medium and severe defect cables condition, the cables condition can
be determined by making holes along the cables. The holes is made by using
nail and the depth of the holes is about ≈ 0.2cm on XLPE. For medium defect
condition, 3 holes are made and for severe defect condition 9 holes are made by
using nail.

5. After that, tank is filled by tap water until it reached half of tank capacity. The
tank is leave for several days for removed chlorine that contain in tap water.

Figure 3.1: Power Cable Immersed in water

6. The experiment setup as shown in Figure 3.1. The cables are placed in tank that
contains tap water and the hole along the cables is immersed in water.

7. The cables are connected in parallel in each other in order to get 2 kV HVDC for
all cables. The cables metal shield is connected to the ground as shown in Figure
3.2.
 16

Figure 3.2: Constructed Experiment Setup

8. 2 kV HVDC voltages were injected to the polymeric cables for 2~3 hours in a
day.

9. After 0, 10, 20, and 40 hours the insulation resistance is measured by using
Megger meter. 2.5kv voltage is injected for 20 second before the readings are
taken. The data is recorded in tables.

10. For measuring electrical parameters. The parameters can be measured by using
Bridge Measurement (Tettex Instruments – Schering Bridge Model 2816). 2 kV
voltage is injected to the cables and the reading of the electrical parameter such
as , capacitance, tan δ and power factor is taken for 0, 10, 20 hours and recorded
in table.

11. Lastly for measuring recovery voltage, recovery voltage meter will be used. The
reading will be taken after 20and 40 hours after the cables ageing.
 17

3.3 Measurement Setup

In order to study the performance and behaviours of underground power cables

due water treeing under HVDC. The measurements of electrical parameters on cables
sample were carried out to determine the insulation resistance, capacitance dissipation
factor (tan δ), power factor and recovery voltage profile of the aged XLPE cables due to
water tree after several hours injection of 2 kV HVDC.

3.3.1 Measurement of Insulation Resistance

The measurements of insulation resistance of the cable were performed by using

advanced digital high voltage insulation tester (Megger meter). Megger meter has two
terminals as shown in Figure 3.3. The red clip function is to inject voltage to cable so it
is connected to the cables conductor. Meanwhile black clip is for grounding (E) purpose
and it is connected to cable metal shielding. The position of conductor and grounding in
cables is shown in Figure 3.4. Then voltage were set to 2.5 kV and time for 20 seconds.
After all the connection were checked and confirmed, the test was performed by
pressing test button (orange button on the left of bottom, Figure 3.3). Three readings
were taken for all sample cables and all data obtained were recorded in Table 1

( Appendix A)
 18

Figure 3.3: Megger Meter

Figure 3.4: XLPE Cables Sample

 19

3.3.2 Measurement of Capacitance, Dissipation Factor (tan δ), and Power Factor

The measurements of capacitance, dissipation factor (tan δ), and power factor
were all performed at frequency 50 Hz using a Tettex Instruments – Schering Bridge
Model 2816 (Appendix C) with automatic guard potential regulator. Bridge is set to
inject 2 kV and, temperatures were set to the room temperature. Room temperature is
been monitored and for every changing in room temperature the bridge temperature
must be set again.

After all the connection were checked and confirmed, the test was performed by
injecting of test voltage ( U ). The values of dissipation factor (tan δ), power factor and
capacitance readings are displayed on the Tettex Instruments – Schering Bridge Model
2816 seven-segment screen and printed as a hard copy. Three readings were taken for
all sample cables and all data obtained were recorded in Table 2, Table 3, and Table 4

( Appendix A )

3.3.3 Measurement of Recovery Voltage Profile

The measurements of voltage recovery profile were all performed by using the
Tettex type 5461 Recovery Voltage Meter. Recovery voltage meter (RVM) determines
the polarization spectra was developed which carries out all measurement automatically
and displays the result in numerical or graphic form. With this method, peak recovery
voltage can be represented as a function of time constants determined from a series of
charge and discharge measurement.
 20

Before proceeding the test, RVM were set to inject 2 kV of voltage. The test
cables has two terminals which are red and black wire. The red wire (warm point) is
connected to conductor during charging and the black one (cold point) is the signal
ground connected to the cables metal shield. RVM is selected to automatic RVM mode.
After all the connections were checked and confirmed, the test was performed by
injecting of test voltage ( U ). The values of voltage recovery profile readings are
displayed on the Tettex type 5461 Recovery Voltage Meter screen and automatically
printed as a hard copy. Recovery voltage profiles were measured for all samples of
cables after 20 and 40 hours ageing process. All the data obtained were recorded in
table ( Appendix B ).
 21

CHAPTER 4

RESULT AND DISCUSSION

4.1 Introduction

In this chapter the result obtained from the experiment will be analysis and
discussed. All the data obtained were recorded in table. For analysis process all data
were plotted in graph. The analysis will focused on the change of the electrical
parameters such as insulation resistance, tan δ, power factor, capacitance, and voltage
recovery profile over the ageing time and ageing level.
 22

4.2 Insulation Resistance

The result of insulation resistance obtained from advanced digital high voltage
insulation tester (Megger meter) were recorded in a table and plotted in graph.

5
Resistance(GΩ)

4
Non-Defect
3
Moderate Defect
2 Severe Defect

0
0 10 20 40
Time(Hour)

Figure 4.1: Resistance vs Time

Figure 4.1 shows the graph of insulation resistance between 3 samples of cables
which are non defect, moderate defect, and severe defect condition for 40 hours ageing
time. The graph show the value of insulation resistance for all sample with difference
ageing level were decrease over the ageing time. The decreasing of insulation resistance
value was influenced by degradation of insulation. Since the cables were immersed in
water, the insulation cables will absorbs the amount of water depend on their ageing
level. Water migration into the insulation and under HVDC stress it leads to form water
trees. The formation of water trees lead to decreased insulation resistance and very
 23

slight increase in leakage current. The relation of leakage current and insulation
resistance can represent in equation 3.

𝑉
𝐼𝑟 𝑙𝑒𝑎𝑘𝑎𝑔𝑒 𝑐𝑢𝑟𝑟𝑒𝑛𝑡 = (3)
𝑅

Where,

Ir = Leakage current

R = Insulation resistance

V = Injected voltage

From the graph above, the severe defect condition cable has the lowest
value of insulation resistance after 40 hours of injection of HVDC compared to non
defect and moderate defect condition cable. This happens due to different level of
ageing. Severe defect condition has a highest ageing level compared with other samples.
Thus water absorption rate into cables insulation faster than other samples. More water
absorbs in sample cables will lead more of formation of water trees and cause more
degradation on insulation cables. However for non defect cable sample the insulation
resistance values higher than moderate defect sample. Possibly this is happen due to
ageing process that already occur in non defect cable sample. The cables samples that
used in this experiment was a used cables that have undergo several test before that.
 24

4.3 Tan δ and Power Factor

The result of dissipation factor (tan δ) and power factor obtained from Tettex
Instruments – Schering Bridge Model 2816 were recorded in a table and plotted in graph

0.004

0.0035

0.003

0.0025
Non Defect
Tan∂

0.002 Medium Defect

0.0015 Severe Defect

0.001

0.0005

0
0 10 20
Time(Hours)

Figure 4.2: Tan δ vs Time (Hours)

The graph of dissipation factor (tan δ) and power factor between 3 samples of
cables which are non defect, moderate defect, and severe defect condition cable were
shown in Figure 4.2 and Figure 4.3. The values of the dissipation factor and power
factor obtained from bridge measurement were approximately same because the values
of tan δ were too small.

tan δ ≈ power factor (4)

 25

0.004

0.0035

0.003
Popwer Factor

0.0025
Non Defect
0.002
Moderate Defect
0.0015 Severe Defect

0.001

0.0005

0
0 10 20
Time(Hour)

Figure 4.3: Power Factor vs Time (Hours)

From the graph as shown in Figure 4.2 and Figure 4.3, they show the value of tan
δ and power factor were increased with ageing time. The increasing value of tan δ and
power factor were affected by degradation of cables insulation due to formation of water
trees. These water trees lead to decreased of insulation resistance and slightly increased
the leakage current ( Ir ). If the leakage current is increased, the value of tan delta will
increase. This will follow the equation of dissipation factor as below:

IR
tan  = (5)
IC
 26

Figure 4.4: Dissipation Factor Model

From graph in Figure 4.2 and Figure 4.3 it also shows the value of tan δ and
power factor for severe defect condition were higher than moderate defect condition.
This can explained by the different of ageing level that lead a degradation in cables
insulation due to water trees that has been explained earlier in insulation resistance
discussion. For non defect condition the value of tan δ and power factor were higher
than severe and moderate cables condition. This result also supports the earlier
hypothesis about the condition of non defect cables sample that had been already
undergoing degradation process.

4.4 Capacitance

The result of capacitance obtained from Tettex Instruments – Schering Bridge

Model 2816 were recorded in a table and plotted in graph.
 27

800

700

600
Capacitance(pF)

500

400 Non Defect

300 Moderate Defect

Seceve Defect
200

100

0
0 10 20

Time(Hours)

Figure 4.5: Capacitance(pF) vs Time(Hours)

From the graph above it shows severe defect condition had a highest capacitance
value followed by moderate defect condition and non defect condition. Over the ageing
time the value of capacitance does not have a significant change, although the
degradation on insulation of the cables samples already occur due to water tress. This
data does not show a correlation between capacitance value and degradation process
because supposedly the value of capacitance should have decrease over the time, proved
by equation 5 and equation 6.

Ic = C dV/dt (6)
 28

This error possibly happen because of mishandle of measurement on Tettex

Instruments – Schering Bridge Model 2816. Other that, there are some error on
capacitance measurement from Tettex Instruments – Schering Bridge Model 2816
because from the graph above the value of capacitance readings does not show a
significant change.

4.5 Recovery Voltage Profile

The result of recovery voltage profile obtained from Tettex instrument

type 5461 Recovery Voltage Meter were recorded in a table and plotted in graph. From
data were taken the recovery voltage profiles were analysis. From the recovery voltage
profile measurement graph the time taken Tc(s) for the voltage to reach peak UrMAX
were compared between difference ageing level and ageing time. Degradation on
insulation will affect the time taken Tc(s) for the voltage to reach peak UrMAX .
Recovery voltage profile were currently use on oil-paper transformer to monitor the
insulation degradation level.

For XLPE insulation there are still in study for using recovery voltage profile to
determine the degradation on XLPE insulation. In this project the time taken Tc(s) for
the voltage to reach peak UrMAX will be studies to find a correlation by using recovery
voltage profile to determine degradation on XLPE insulation.
 29

1.4

1.2 Peak
Voltage
1

0.8
Urmax(V)

0.6 20 Hour
40 Hour
0.4

0.2

TC(s)

Figure 4.6: Recovery Voltage( Non Defect)

Figure 4.6 shows a recovery voltage profile on non defect condition for
20 and 40 hours ageing time. It shows for 20 hours ageing time, the time taken Tc(s) for
the voltage to reach peak UrMAX is 10 second and for 40 hours ageing time, time taken
Tc(s) for the voltage to reach peak UrMAX is 5 second. It shows the more time taken to
ageing, the time taken Tc(s) for the voltage to reach peak UrMAX become more faster.

From earlier analysis and discussion, the degradation process increasing by

ageing time. From the recovery voltage profile result it shows the more degradation on
insulation, the less time taken Tc(s) for the voltage to reach peak UrMAX . The recovery
voltage profile result also give the similar pattern for recovery voltage profile on
moderate and severe defect condition as shown in figure 4.7 and Figure 4.8.
 30

5 Peak
Voltage
4

3
URMAX(v)

20 Hour
2 40 Hour

-1

-2
Tc(s)

Figure 4.7: Recovery Voltage ( Moderate Defect)

14

12
Peak
10 Voltage

8
URMAX

6 20 Hour

4 40 Hour

-2

TC(s)

Figure 4.8: Recovery Voltage (Severe Defect)

 31

7 Peak Voltage

5
URMAX

4 Non Defect
3 Moderate Defect

2 Severe Defect

-1

-2

Tc(s)

Figure 4.9: Recovery Voltage(20 Hours)

14

12
Peak
10
Voltage

8
Urmax

6 Non Defect
Moderate Defect
4 Severe Defect

-2

Tc(s)

Figure 4.10: Recovery Voltage( 40 Hours)

 32

Figure 4.9 shows a recovery voltage profile for 20 hours ageing time on different
ageing level. Severe defect condition of cables sample shows a fastest time taken Tc(s)
for the voltage to reach peak UrMAX follow by moderate defect and non defect
condition. Also for 40 hours of ageing time as shown in Figure 4.10, time taken Tc(s)
for the voltage to reach peak UrMAX for severe defect condition is more fastest than
moderate defect condition and follow by non defect condition.

From earlier discussion, since in severe defect condition the water absorption
rate into cables insulation is much faster than other samples. More water absorb in
sample cables will lead more of formation of water trees and cause more degradation on
insulation cables. It shows the higher degradation of insulation cause by water trees, the
fastest time taken Tc(s) for the voltage to reach peak UrMAX.
 33

CHAPTER 5

CONCLUSION AND RECOMMENDATION

5.1 Conclusion

Water trees are an important factor that can influence the insulation
electrical parameters such as insulation resistance, capacitance, power factor, tan
δ and recovery voltage profile. The values of insulation resistance is more lower
for severe defect condition than moderate defect condition.

Tan δ and power factor shows severe defect condition have higher
readings value than moderate defect condition. For capacitance, the increasing
of defect level on insulation give higher reading value of capacitance. Recovery
voltage profile shows a correlation between ageing time and ageing level. More
time taken for ageing, less time taken for voltage reach peak value.
 34

Higher ageing level gives shortest time for voltage to reach peak value.
From insulation resistance, tan δ, power factor and recovery voltage profile
analysis, it show an evidence of degradation process has already occur on non
defect cable condition before the experiment.

5.2 Recommendation

To ensure a better research in future, there are few suggestions that can be as a
guideline for improving the experiment and measurement setup to get more accurate
results. First, it is recommended to used a new cables and the condition is checked first
to ensure the new cables is perfectly heath before proceeding to experiment setup. Other
recommendation is to continuously doing ageing process by continuously injecting
HVDC on samples. This will result of maintaining the formation of water trees in XLPE
insulation.

Furthermore, it is really recommended to measure the leakage current value for

all cables sample. It can help to improve analysis process on other electrical parameter
such as insulation resistance, dissipation factor, power factor and capacitance.
 35

REFERENCES

1. Ian Clarke (2005). Revitalizing water Damages Cables from Within. Trenchless
Technology International.

2. J. Castellon, S. Agnell, A. Toureille I, G. Platbrood ,Y. Tits, P. Adriaensens , R.

Carleer and D. Vangulic (2005). A New Diagnosis of the XLPE Insulation with
Water Treeing by Space Charge Measurements with the Thermal Step Method
and Water Nuclear Magnetic Resonance Imaging. Annual Report Conference on
Electrical Insulation and Dielectric Phenomena : IEEE .

3. Electrical Treeing. http://en.wikipedia.org/wiki/Electrical_treeing.

4. Sensor Technology.
http://www.ee.washington.edu/research/seal/projects/seal_robot/sensors.html.

5. Steven Boggs, James Xu (2001). Water Treeing Filled versus Unfilled Cable
Insulation. IEEE Electrical Insulation Magazine. Vol. 17, No.1.

6. Vern Buchholz. Finding the Root Cause of Power Cable Failure.

http://www.electricenergyonline.com.

7. M. Haddad and D. Warne (2004). Advances in High Voltage Engineering. The

Institution of Electrical Engineer, London, United Kingdom.
 36

8. M.I. Qureshi, N.H. Malik; A.A. Al-Arainy, M.N. Saati, O.A. Al-Nather, and S.
Anam (2008). Investigation of Resistance Toward Water Tree Growth in XLPE
Insulated MV Cables Produced in the Arabian Gulf Region. 2008 International
Conference on Condition Monitoring and Diagnosis. April 21-24. Beijing,
China: IEEE.

9. William R. Stagi (2006). Cable Injection Technology. Proceedings of the 2006

IEEE Latin America Conference : IEEE.

10. Tom A. Short (2006). Electric Power Distribution Equipment and Systems.
CRC Press Taylor & Francis Group.

11. High-voltage direct current. http://en.wikipedia.org/wiki/High-

voltage_direct_current.

12. Hakan Lennartsson. Modern XLPE Materials for Extruded Energy Cable
Systems. Power Cables & Switchgear Forum – IDC Technology.
 37

APPENDIX A

Electrical Parameter Result

Table 1: Resistances (GΩ )

Cables Non-defect Medium-defect Severe-defect

Hours Readings average Readings average Readings average

2.13 5.5 1.45

0 2.18 2.17 5.7 5.77 1.51 1.53

2.19 6.1 1.63

1.36 2.95 1.26

10 1.18 1.24 3.64 3.47 1.29 1.29

1.18 3.84 1.32

1.20 3.16 1.19

20 1.23 1.22 3.45 3.41 1.22 1.21

1.22 3.63 1.22

1.23 2.98 1.10

40 1.19 1.19 3.43 3.21 1.09 1.13

1.16 3.22 1.20

 38

Table 2: Tan δ

Cables Non-defect Medium-defect Severe-defect

Hours Readings average Readings average Readings average

0.0003 0.0005 0.0007

0 0.0004 0.0004 0.0004 0.0004 0.0006 0.0006

0.0004 0.0004 0.0006

0.0018 0.0002 0.0006

10 0.0013 0.0015 0.0002 0.0002 0.0005 0.0005

0.0013 0.0002 0.0005

0.0038 0.0007 0.0007

20 0.0032 0.0035 0.0005 0.0006 0.0005 0.0006

0.0034 0.0005 0.0005

 39

Table 3: Power Factor

Cables Non-defect Medium-defect Severe-defect

Hours Readings average Readings average Readings average

0.0002 0.0005 0.0007

0 0.0004 0.0003 0.0004 0.0004 0.0006 0.0006

0.0004 0.0004 0.0006

0.0018 0.0002 0.0006

10 0.0014 0.0015 0.0002 0.0002 0.0005 0.0005

0.0013 0.0002 0.0005

0.0038 0.0007 0.0007

20 0.0032 0.0035 0.0005 0.0006 0.0005 0.0006

0.0034 0.0005 0.0005

 40

Table 4: Capacitance (pF)

Cables Non-defect Medium-defect Severe-defect

Hours Readings average Readings average Readings average

602.6 661.7 732.3

0 602.4 602.5 661.7 661.7 732.3 732.3

602.4 661.7 732.4

602.0 660.0 730.8

10 602.1 602.1 660.0 660.0 730.7 730.7

602.1 660.0 730.7

601.8 659.9 730.8

20 601.6 601.7 659.6 659.7 730.2 730.3

601.6 659.6 729.9

 41

APPENDIX B

Recovery Voltage Profile Results

Table 5: Cable 1 ( Severe Defect Condition ) 20 Hours

20 hours ageing period

Cycle Tc (s) Td (s) Urmax (V) dUr/dt (V/s) Tpeak (s)
1 0.02 0.01 3.98 83.51 3
2 0.05 0.025 4.11 62.91 0.5
3 0.1 0.05 4.081 40.01 0.9
4 0.2 0.1 4.334 20.13 1.2
5 0.5 0.25 6.439 15.6 1.2
6 1 0.5 7.596 16.36 1.4
7 2 1 5.467 9.084 2.2
8 5 2.5 2.273 1.186 6.5
9 10 5 1.952 0.584 15.2
10 20 10 1.912 0.107 24.2
11 50 25 2.003 -0.091 31.8
12 100 50 1.995 -0.089 61.4
13 200 100 2.054 0.027 107.0
14 500 250 2.116 0.172 82.2
15 1000 500 2.482 0.137 80.8
16 2000
 42

Table 6: Cable 2 ( Medium Defect Condition ) 20 Hours

20 hours ageing period

Cycle Tc (s) Td (s) Urmax (V) dUr/dt (V/s) Tpeak (s)
1 0.02 0.01 5.999 135.8 0.1
2 0.05 0.025 2.936 75.96 0.1
3 0.1 0.05 -0.139 6.947 0.1
4 0.2 0.1 -1.139 -8.861 0.1
5 0.5 0.25 -1.343 -9.946 0.1
6 1 0.5 -1.016 -8.261 0.1
7 2 1 -0.385 -3.805 0.1
8 5 2.5 1.086 -0.313 13.9
9 10 5 1.237 0.259 13.1
10 20 10 1.361 0.217 15.6
11 50 25 1.432 0.126 18.5
12 100 50 1.677 0.027 27.2
13 200 100 1.878 0.027 32.8
14 500 250 2.088 0.120 43.0
15 1000 500 2.195 0.131 45.2
16 2000
 43

Table 7: Cable 3 ( Non Defect Condition ) 20 Hours

20 hours ageing period

Cycle Tc (s) Td (s) Urmax (V) dUr/dt (V/s) Tpeak (s)
1 0.02 0.01 1.084 35.98 0.1
2 0.05 0.025 0.132 0.338 3,5
3 0.1 0.05 0.214 0.677 4.0
4 0.2 0.1 0.302 0.678 4.5
5 0.5 0.25 0.414 0.668 4.9
6 1 0.5 0.514 0.632 5.0
7 2 1 0.613 0.533 4.7
8 5 2.5 0.755 1.714 2.9
9 10 5 0.854 3.257 1.4
10 20 10 0.836 2.988 1.8
11 50 25 0.810 1.860 4.4
12 100 50 0.815 0.621 8.0
13 200 100 0.872 0.532 9.9
14 500 250 0.954 0.349 14.0
15 1000 500 0.996 0.208 23.0
16 2000
 44

Table 8: Cable 1 (Severe Defect Condition) 40 Hours

40 hours ageing period

Cycle Tc (s) Td (s) Urmax (V) dUr/dt (V/s) Tpeak (s)
1 0.02 0.01 4.261 26.09 0.9
2 0.05 0.025 9.423 64.44 0.8
3 0.1 0.05 12.11 89.34 0.7
4 0.2 0.1 9.822 72.26 0.6
5 0.5 0.25 1.695 16.97 0.3
6 1 0.5 -0.153 -6.24 0.1
7 2 1 -0.293 -9.395 0.1
8 5 2.5 -0.257 -8.877 0.1
9 10 5 -0.121 -4.173 0.1
10 20 10 -0.024 -0.936 0.1
11 50 25 1.852 0.048 66.1
12 100 50 1.922 0.078 74.2
13 200 100 1.765 0.065 78.2
14 500 250 1.760 0.042 103.2
15 1000 500 1.608 0.034 112.9
16 2000
 45

Table 9: Cable 2 (Medium Defect Condition) 40 Hours

40 hours ageing period

Cycle Tc (s) Td (s) Urmax (V) dUr/dt (V/s) Tpeak (s)
1 0.02 0.01 0.882 5.794 3.6
2 0.05 0.025 2.523 9.545 2.3
3 0.1 0.05 3.800 12.29 1.7
4 0.2 0.1 5.294 18.88 1.1
5 0.5 0.25 2.513 11.52 0.6
6 1 0.5 -0.035 -4.166 0.1
7 2 1 -0.213 -8.489 0.1
8 5 2.5 -0.283 -11.067 0.1
9 10 5 -0.173 -7.274 0.1
10 20 10 -0.055 -2.412 0.1
11 50 25 -0.005 -0.117 0.1
12 100 50 2.064 0.039 141.4
13 200 100 2.021 0.034 189.6
14 500 250 1.747 0.032 165.2
15 1000 500
16 2000
 46

Table 10: Cable 3 (Non Defect Condition) 40 Hours

40 hours ageing period

Cycle Tc (s) Td (s) Urmax (V) dUr/dt (V/s) Tpeak (s)
1 0.02 0.01 0.696 29.93 0.1
2 0.05 0.025 0.257 0.092 4.3
3 0.1 0.05 0.466 0.688 3.6
4 0.2 0.1 0.627 1.102 3.7
5 0.5 0.25 0.819 1.274 4.0
6 1 0.5 0.977 1.234 3.8
7 2 1 1.121 0.964 3.4
8 5 2.5 1.270 3.458 2.1
9 10 5 1.264 4.681 1.5
10 20 10 1.153 4.483 1.6
11 50 25 1.017 3.354 3.1
12 100 50 1.011 2.118 4.6
13 200 100 1.010 1.275 9.3
14 500 250 0.966 0.476 20.1
15 1000 500 0.913 0.314 22.7
16 2000
 47

APPENDIX C

Measurement Setup

Figure 5: Tettex Instruments – Schering Bridge Model 2816

 48

Figure 6: Sample connection to Tettex Instruments – Schering Bridge Model 2816

 49

Figure 7: Tettex instrument type 5461 Recovery Voltage Meter