torelli za ucinke

8/10/2019 Torelli Za Ucinke

1/180

1/138

Transient Overvoltages and Distance Protection:Problems and Solutions

A thesis submitted in fulfillment of the requirements for the

degree of Master of Engineering by Research

Leonardo Torelli

B.Eng. (Hons)

School of Electrical and Computer Engineering

RMIT University

November 2010


2/180

2/138

Table of Contents

The Author .......................................................................................................................... 5

Acknowledgement...............................................................................................................6

Declaration .......................................................................................................................... 7

Abbreviations ...................................................................................................................... 8

Executive Summary ..........................................................................................................10

1 Introduction .................................................................................................................. 11

2 Neutral Earth Resistor and Transient Overvoltages .....................................................14

2.1 Introduction ..........................................................................................................14

2.2 Method of Earthing............................................................................................... 15

2.3 Review of the Existing Body of Knowledge ........................................................16

2.4 Design of the Experiment .................................................................................... 25

2.4.1 The Software Package .................................................................................. 252.4.2 The Model .................................................................................................... 26

2.4.3 The experimental design............................................................................... 27

2.4.4 System Configuration ................................................................................... 29

2.4.5 Simulation Characteristics ............................................................................ 31

2.4.6 Model Components ...................................................................................... 32

2.5 Simulation Results................................................................................................40

2.5.1 Simulation 1 : Unloaded Line - Fault at the End of the Line ....................... 41

2.5.2 Simulation 2 : Unloaded Line - Fault at the Beginning of the Line ............. 47

2.5.3 Simulation 3 : Capacitor Bank. Fault at the End of the Line....................... 50

2.5.4 Simulation 4 : Capacitor Bank. Fault at the Beginning of the Line ............. 56

2.5.5 Simulation 5 : Light inductive Load. Fault at the End of the Line............... 59

2.5.6 Simulation 6 : Energized 66/22kV Transformer Fault at the 66kV Bus. .... 62

2.5.7 Overvoltages and Traveling Waves.............................................................. 63

2.5.8 Impact of Time of Fault with the Overvoltage ............................................. 68

2.5.9 Overvoltage Response on Phase a ,b and c .................................................. 72

2.5.10 Overvoltage Harmonic Response ................................................................. 76

2.6 Conclusion............................................................................................................78

3 CVT and Transient Overvoltages ................................................................................. 80

3.1 Overview .............................................................................................................. 80


3/180

3/138

3.2 The CVT Model ..................................................................................................81

3.3 Simulation Settings............................................................................................... 843.4 Simulation Results................................................................................................86

3.5 CVT Conclusion .................................................................................................100

4 Distance Protection Scheme ....................................................................................... 101

4.1 Introduction ........................................................................................................101

4.2 MHO Characteristics Theory ............................................................................. 107

5 Transient Overvoltages and Distance Protection: Solutions ...................................... 111

5.1 Introduction ........................................................................................................111

5.2 Polarisation Techniques...................................................................................... 113

5.3 Polarising Techniques ........................................................................................ 117

5.4 CVT Techniques.................................................................................................119

5.5 Setting Advice ....................................................................................................121

6 Conclusion..................................................................................................................125

6.1 Research Findings ..............................................................................................125

6.2 Areas for Further Investigation........................................................................... 127

7 References .................................................................................................................. 129

Appendix A ..................................................................................................................... 133

10th International Conference on Developments in Power System Protection, March

2010, Manchester, UK -Abstract.....................................................................................133

Appendix B ..................................................................................................................... 134

Overvoltage StudyPSCAD Data .................................................................................. 134

Appendix C ..................................................................................................................... 135

Overvoltage studySteady State Voltage Data ............................................................. 135

Appendix D ..................................................................................................................... 136

Overvoltage Study- PSCAD Model ................................................................................ 136

Appendix E......................................................................................................................137

Voltage and Current as function of SIR.......................................................................... 137


4/180

4/138

Appendix F......................................................................................................................138

CVT Study

PSCAD Data .............................................................................................138


5/180

5/138

The Author

My name is Leonardo Torelli. I have been working for fourteen years in power

industry. I have worked in Italy in Utilities and in Australia in consulting firms. I am

currently working at Hydro Tasmania Consulting and involved in Protection and

Power System Analysis projects.

It is not a surprise that the topic of this research goes across these two areas of

electrical engineering. The aim of this study was, indeed, proving how important is

the knowledge of power system analysis for Protection Engineers.

As result of this study, I presented a paper at the 10th International Conference on

Developments in Power System Protection, March 2010, Manchester, UK

This research has given me the opportunity to expand my skills, challenging my

knowledge and improved my analytical skills. I am committed to use this

experience at RMIT University as starting point for further studies.


6/180

6/138

Acknowledgement

I thank Dr Selva Moorthy for his contribution, support and expert advice during this

study. I have appreciated his sincere interest in my professional development and

support of my personal life. Dr Moorthys feedback helped me to stay in line with

my schedule and achieve the completion of this research.

I would like to thank my former and existing colleagues at Hydro TasmaniaConsulting for their encouragement during this journey.

Last, I would like to thank my three beautiful children, my family and friends for

their warm and enthusiastic appreciation of my work.

.


7/180

7/138

Declaration

I certify that except where due acknowledgement has been made, the work is that

of the author alone; the work has not been submitted previously, in whole or in

part, to qualify for any other academic award; the content of the thesis is the result

of work which has been carried out since the official commencement date of the

approved research program; and, any editorial work, paid or unpaid, carried out by

a third party is acknowledged.

.

Leonardo Torelli

Date


8/180

8/138

Abbreviations

The following terms are used in this thesis:

! HVDC - High Voltage Direct Current

! MHO relay - Distance relay with circular characteristic which passes through

the origin of the R-X plane

! NER - Neutral Earth Resistor

! NEX - Neutral Earth Reactance

! PSCAD-EMTDC - Power system software package

! SIR- Source Impedance Ratio

! TRV- Transient Recovery Voltage


9/180


10/180


11/180

11/138

1 Introduction

Power systems are designed and operated to supply electrical energy to the

customers in a safe, reliable and economical way. Faults represent one of the main

challenges for the power system. Among all the types of faults, phase to ground

faults represent the majority of the events, with records between 80 to 90 % of all

faults. Ground faults are generated from insulation breakdown, atmospheric

conditions and accidental contacts of birds or branches of trees with power lines.

Therefore, these faults are transitory in nature.

During this post fault initial period, the voltage change dramatically from a pre fault

steady state value to a post fault steady state value. In addition, the voltage signal

is disturbed by higher and lower frequency components. Therefore, transient

overvoltages could affect the accuracy of the protection scheme. At the same time,

the protection scheme should be able to operate in a fast manner to reduce the risk

to personnel, equipment damage, and system stability.

It is evident that achieving a short operating time and high accuracy of the

protection scheme represents a major challenge for protection relay manufacturers

and protection engineers. This research investigated the performance of modern

digital distance relay during the transient overvoltages period which follows a

ground fault.

Initially, this study focused on power system analysis to have a clear understanding

of the transient overvoltage phenomena. The overvoltage study was conducted

using PSCAD-EMTC software package. The investigation was performed by using

the experimental method that involves the scientific manipulation of the variables

involved in the process and the systematic study of the behavior of the system.

The study was conducted on a radial system using a neutral earth resistor, NER,

as a method to earth the 66 kV network. However, the theory and the results


12/180

12/138

obtained are largely applicable to higher transmission voltages and in power

systems earthed via a neutral earth reactor, NEX.

Following the transient overvoltages study, the research focused on the

performance of five modern digital distance relays.

In summary, the project objectives are as follow:

! Determine the characteristics of transient overvoltages generated by a

phase to ground fault. This work should highlight the main factors that

influence the voltage disturbances

! Analyze the characteristics of modern digital distance relays to determine if

the transient overvoltages could affect the accuracy ad operating time of

the protection relay

! Establish a method to predict the performance of the distance protection

relay following a phase to ground fault

! Elaborate a list of setting advice that could enhance the application of the

distance protection relay in relation to the transient overvoltages

disturbance

This study aimed to validate the existing body of knowledge, highlight new insight

and views and determine new findings.

The research is organised in six Sections:

! Section 2 -NER and Overvoltage

! Section 3 - NER and Capacitor Voltage Transformers

! Section 4- NER and Distance Protection

! Section 5- Distance Relay Applications


13/180

13/138

In the last part of this study, Section 6, research conclusions and recommendations

for areas of further investigations are also provided.


14/180


15/180

15/138

2.2 Method of Earthing

The earthing of the power system involves the implementation of an electrical

connection between the neutral point and the ground.

Connection of the neutral to the earth can be done in several methods:

! Solidly earthing

! Neutral earth resistor, NER

! Neutral earth reactance, NEX

! Compensated earthing

! Ungrounded

The type and size of the earthing system will affect the following system

parameters [1-4]:

! Fault current

! Personnel safety

! System Protection

! Steady state overvoltage during ground faults

! Transient overvoltage during ground faults

! Thermal stress on the equipment

! Effect on communication circuit

! Harmonic current on neutral connection

! System stability


16/180


17/180

17/138

Transient overvoltages are produced by sudden changes in the electrical system

as caused by lightning, faults or switching [6]. The method of earthing system doesnot have any effect on the system during normal operation [4]. However, during the

transient period the method of earthing has an impact on the system response and

magnitude of the transient overvoltage.

Power demand, power generated and system configuration vary continuously with

time. Therefore the power system always changes from one steady state to

another steady state. In this very short periods current and voltages may reach

dangerous values for the equipment and the insulation of the system.

Overvoltages are often classified according to their duration in two groups [7, 8]:

! Temporary Overvoltage above 200 ms

! Transient Overvoltage below 200 ms

Alternatively, overvoltages can be classified according to their origin:

! External overvoltage caused by atmospheric phenomena as lighting and

electrostatic charges

! Internal overvoltage generated within the power systems

This type of disturbances are also classified according to the frequency of the two

major overvoltage components [9]:

! Power frequency overvoltage

! Natural frequency overvoltage of a short duration superimposed on the

power frequency overvoltage

The sum of the power and natural frequency overvoltage is the voltage recorded in

the field. This voltage is commonly defined as a transient overvoltage.


18/180

18/138

The overvoltage component with a power frequency does not theoretically have

any decaying component and it often termed as steady state overvoltage. Instead,the short duration component usually decays in less than 100 ms. Above all; the

magnitude of the overall overvoltage is of greatest interest for the impact on the

planning, design and financial cost of the power system.

A fault in the power system can be represented and analysed as the closing of a

switch in the electrical system. This change in the power system develops a new

redistribution of stored energy in the system [6]. Phase to ground faults represent

the majority of faults in the power system. These types of faults are produced by

atmospheric condition, mechanical breakdown of the insulation, objects such as

birds or branches or trees in contact with the overhead line and poles or structures.

Recent statistics conducted by protection relay manufacturers found that these

types of faults represent between 80 and 90 % of the faults in a given power

system [3, 4, 10]. A large majority of these faults involving overhead lines also

have transient characteristics and, therefore, self extinguish in a short period of

time.

When a phase to ground fault occurs, an overvoltage can be measured in the

system during the fault itself or after clearing [10]. The maximum transient

overvoltage is obtained by adding the peak of the temporary overvoltage at power

frequency overvoltage at 50Hz sinusoidal pattern to the transient overvoltage

which usually has a higher frequency. This relation is based on the worst case

scenario with the assumptions that the two components will have maximum peak

values at the same time. This is plausible because of the different time frames.Power frequency peak lasts long in the context of high frequency oscillations.

The magnitude of the power frequency overvoltage is strictly related to the

characteristic parameters of the system. This overvoltage can be calculated using

the symmetrical components theory. Considering some system simplification as

necessary, this engineering technique produces accurate results. The fundamental

frequency voltage, which is also referred to the steady state overvoltage, is absent


19/180

19/138

in a solidly earthed system. In an ungrounded system the overvoltage reaches the

maximum value of !3=1.732. In this latter case, the phase to ground voltage isequal to the phase to phase voltage.

Application of the neutral earth resistor as a method of grounding increases the

fundamental frequency overvoltage from none to a maximum of 1.732 for the

insulated system.

Between these two boundaries of the overvoltage measured in the system, the

institution of Electrical and Electronic Engineers established an arbitrary limit of

the overvoltage measured, which is named effective grounding[11].

The effective grounding term indicates that, during a ground fault, the voltage on

the healthy phases will not exceed 80 % of the maximum line to line voltage. The

term effective grounding is commonly used in the power industry as a point of

reference for the design of the earthing system and sizing of the neutral

impedance. Therefore a power system can be effectively grounded or

ineffectively grounded. In other words, in a 66kV subtransmission system, thephase to ground voltage during fault condition will be:


20/180

20/138

V Nominal

Vph-ph=66kV

Vph-g = 66kV/!3=38.1 kV

Solidly Grounded System

Vph-g fault= 66/ !3 = 38.1 kV

Effective grounding System

Vph-g fault= 66 * 0.8= 52.8 kV

Ungrounded System

Vph-g fault= 66kV

The main advantage of having an effective grounded system is limiting the

overvoltage magnitude which has a direct impact on the cost of the system. For

instance, the required insulation medium can be reduced with direct benefits to the

cost of the electrical equipment. This advantage is a key factor for the design of

subtransmission and transmission power systems.

The effective grounding system is nowadays considered the best compromise

between reducing the phase to ground fault and keeping the overvoltages at a

reasonable level. Therefore the effective grounding is the preferred solution for the


21/180

21/138

higher voltages in countries where the power system is grounded via an

impedance[4] .

How to achieve an effective grounded system? IEEE Recommended Practice for

Grounding of Industrial and Commercial Power Systems states the required

system conditions:

! The system must be grounded through a sufficient low impedance such

that for all system conditions the ratio of zero sequence reactance to

positive sequence reactance is positive and equal or less than ( X0/X1"3)

and

! the ratio of zero sequence resistance to positive-sequence reactance is

positive and less than 1( R0/X1


22/180

22/138

! 20

1

0"

X

X, which is inherently met in resistance grounding and with delta

star power transformers [12], unless the power system is earthed via a

large reactor

From the system conditions stated above, it can be derived that the size of neutral

earth resistor or neutral earth reactor is not critical by itself for the magnitude of the

overvoltage. The relation with other system parameters determines the magnitude

of the overvoltage. The above conditions also highlights that unplanned increase of

the size of the resistor will increase the value of the overvoltage due to the fault

clearing restrike.

Besides, the system conditions show that the reactance of the overall zero

sequence network shall be inductive and not capacitive.

These system conditions aim to limit transient overvoltages and, at the same time,

reduce the phase to ground fault current between 10 and 25 % of the steady state

fault current [9]. The lower level is dictated by the minimum current requirement for

relay operation. Nowadays with modern digital relays, this requirement could be

reconsidered. The 25 % upper level takes into consideration the resistor cost in

comparison to the reactive option.

Plots of power frequency overvoltage following faults as a function of different

system parameters or system parameters or ratios of system parameters have

been published and are available in the existing literature[9].

Existing studies on overvoltage and impact of the neutral earth resistor on the

measured overvoltages are based on field testing, software simulation and

theoretical analysis.

First researches on transient overvoltages during faults were conducted before the

Second World War. Overvoltages were analytically determined and tested using


23/180

23/138

miniature models[9] . Maximum peak voltages were recorded mainly between 3

and 4 pu with 20 % above 4 pu. The researchers also considered current choppingproduced by the circuit breakers and restrike phenomena known as transient

recovery voltage, usually abbreviated as TRV. Overvoltage calculations and tests

using miniature system have also been conducted by other researchers[6, 8].

These papers provide a complete explanation of the maximum overvoltage in

relation to system parameters and formed the basis for future IEEE standards.

Measurements of overvoltages following a fault were performed in the distribution

network in the 10kV Croatian distribution network[13]. The system was earthed

using the NER. Maximum overvoltage factors are below 2 pu and minimum

variations are recorded for different values of the resistor. A transient software

program was used to validate the results.

It is possible to determine the maximum transient overvoltage by applying the RLC

circuit analysis and the use of symmetrical components. [14]. Generalized plots

which took into consideration the damping effect of the resistor were also

produced.

An overall indication of the effect of the resistor is defined by:

020 XR " circuit is oscillatory

020 XR $ critical damped

020 XR # over damped

Recent studies usually apply the traveling waves theory. In line with this theory,

modern transient software packages also avoid the use of lumped elements [15].

Lumped elements representation is adequate for steady state analysis but give

inaccurate results in transient studies. In the latter studies, where travel time of the

electromagnetic waves and the energy exchange between capacitance and

inductance must be taken into consideration, traveling waves theory is the most


24/180

24/138

accurate method for the computation of the overvoltages along the transmission

line.

Fault generated traveling waves propagates not only on the overhead line to line.

but also line to earth with different propagation parameters [16]. A phase to ground

fault can be represented as disturbance that generates current and voltage

traveling waves on a subtransmission line. At each point of discontinuity the

traveling waves will be refracted and reflected according to the surge impedance

characteristic of the line [17].It is possible to represent the large number of forward

and backwards waves and determine the maximum overvoltage generated as a

function of time by using the Bewley Lattice diagram. However, there are also

some other considerations which must be dealt with which include mutual coupling

with other conductors and wave shape distortion experienced along the

transmission line.

It is important to point out that overvoltage studies and traveling wave application

mainly focus on switching and lightning transients [16]. However, overvoltages

produced by switching episode can be used as a reference for this type of study.

To conclude, what overvoltage magnitude can be developed under a ground fault?

From the existing literature it can be inferred generally that the maximum

overvoltages following a phase to ground fault are below 3 pu [18-20]. Maximum

overvoltage can instead reach 4.1 pu with trapped charges on the line as in the

case of switching a capacitor bank connected on the line [19].


25/180

25/138

2.4 Design of the Experiment

2.4.1 The Software Package

Usually, software packages are used to facilitate transient overvoltage studies. In

fact, these investigations involve complex and time consuming computation that

can be achieved with sufficient accuracy with software packages.

PSCAD-EMTDC is a state of the art software package developed using the wave

propagation theory first published by DR Hernmann W Dommel in the transaction

of IEEE Power Apparatus and Systems in 1969. based on this paper another two

transient analysis software packages well known in the industry, EMTP and, later,

the ATP software package [21] were developed.

PSCAD is a very powerful and flexible graphical interface that uses the EMTDC

simulation program. The first version of PSCAD was produced by Denis Woodford

to model and simulate the high voltage DC system in Canada. The software is now

developed and maintained by the Manitoba HVDSC Research Centre and is

continually upgraded to incorporate the latest validated research.

The software allows one to model an electrical circuit, run a simulation, analyze the

results and manage the data information in an efficient way. The software package

is provided with an equipment library, but also allows the creation of new models

and building a personal equipment library. It is used by consulting engineers,equipment manufacturers and laboratories for planning and design of power

systems.

Typical projects involve transient stability, dynamic stability, relay coordination,

transformer saturation, insulation coordination studies, HVDC studies, harmonic

studies, power electronics studies and optimisation of controller parameters.

Further information is located on the website http://www.pscad.com
http://www.pscad.com/


26/180


27/180

27/138

Steady state overvoltage for a given value of the neutral earth resistor was

calculated by applying symmetrical components theory. Factor K2 and K0 wereused to determine fault currents and steady state overvoltages[4].

Z2/Z1K2 $

Z0/Z1K0 $

Where:

Z1= Positive system impedance

Z2= Negative system impedance

Z0= Zero system impedance

The factor K0 gives us a prompt understanding of the power frequency overvoltage

on the healthy phases during a ground fault.

The factor K2 is assumed to be 1. If the fault occurs close to the generator the

factor K2 may increase to 1.4 [4].

Knowing the steady state overvoltage and the total overvoltage, the transient

component can be calculated with:

Natural Frequency V = Measured Overvoltage - Power Frequency V

2.4.3 The experimental design

The most relevant factors that influence the overvoltage magnitude during phase to

ground faults are as follow:


28/180

28/138

! Method of earthing

! Type of fault

! System configuration

! Number of earthing points and location in respect to the fault location

! Electrical parameters of the equipment

! Size of the neutral earth resistor

! Location of the fault

! Time of the fault applied

! Type of subtransmission network as underground cables and overhead

lines have different electrical characteristics

! Type of overhead line

! Length of the line

! Fault impedance

Each variable was analyzed in order to determine the most critical and relevant

factors to be used in this research. In the end, the study was based on thefollowing five factors:

! System configuration

! Type of faults

! Size of the neutral earth resistor


29/180

29/138

! Location of the fault

! Time of the fault

2.4.4 System Configuration

Four system configurations were selected. These configurations were considered

the most likely configurations to produce the highest overvoltages during a ground

fault [5, 14]:

66kV Subtransmission Line

T

#1 #2RL

Resis

+

Figure 2.1: Overhead line with no load at the end of the line model

66kV Subtransmission Line

T

#1 #2RL

Resis

+30 [MVAR]

Figure 2.2: Overhead line with capacitor bank at the end of the line


30/180


31/180

31/138

Faults were simulated at the beginning and at the end of line.

Time of fault

Considering the symmetry of the alternating voltage cycle, the samples were

selected only on one half cycle. As one semi cycle at 50Hz takes 10 ms, 21 sample

spaced 0.5 ms apart were selected.

Line length

Line lengths were set to 100, 10 and 0.2 km.

These line lengths are based on the minimum, maximum and typical line length in

the Victorian Network. Typical subtransmission lines vary between 10-20 km in the

urban area around Melbourne and increase to 30-40 km in the country area. In

Victoria the maximum line length is about 80 Km. The line length of 100 km reflects

the need to investigate the system in the worst case scenario.

2.4.5 Simulation Characteristics

In this study no randomization factor is applied as the chance of biasing the results

in a software simulation is considered negligible. Therefore, simulation studies

were conducted in a logical order to maximize the efficiency of the work.

The simulations were conducted using PSCAD multiple run function. This software

tool allows completion of complex studies in an efficient manner by manipulating

variables sequentially and automatically. PSCAD Multiple Run tool can run and

modulate up to six variables.


32/180

32/138

Maximum instantaneous voltage peak was recorded for each run and on each

phase. Only the maximum overvoltage among the three phases was selected forthe maximum overvoltage record.

Duration of the run time was set to 0.6 s with fault start at 0.4 s plus selected time

delay. The end time of the fault was 0.1 s with clearing fault at the first zero

crossing intersection of the fault current.

According to the Nyquist_Shannon Sampling Theorem, to avoid aliasing and

incorrect representation of the selected variable, sampling rate should be at least

double of the variable frequency to be represented. PSCAD recommends that 50s

for the time step of the process and computation would be sufficient. However,

time step resolution was reduced to 20 s to capture transient components up to

25 kHz [22].

Sampling rate=20 s

Frequency rate= 1/T=106/20= 50kHz

2.4.6 Model Components

Source

The source represents the power system which is upstream the 220/66 kV power

transformer in the substation. The voltage source selected from the PSCAD library

enabled to specify the positive, negative and zero-sequence impedance.

The impedance value is selected to produce a three fault current of 30 kA on the

220 kV bus at the substation. Positive impedance is equal to negative impedance.

Zero impedance was set to be equal to the positive sequence impedance to allow

for a phase to ground fault level equal to the three phase fault level.

The X/R ratio was set to 10.


33/180

33/138

The control system for the output voltage was set using the Fixed Controlfunction

offered by PSCAD software package. Through several trial and error experimentsthis type of control produced a model behavior very similar to an infinite bus. This

setting enhances the robustness of the source model in respect to disturbances on

a lower voltage level. It is assumed that the impact of the overvoltage on the 66 kV

radial line is limited on the higher transmission network.

In summary:

Source Type = Three Phase Voltage-Source Model 3

MVA = 100 MVA

Vph- Vph = 220 kV

Frequency = 50 Hz

Time constant = 0.06 s

Z1=Z2=Z0= 4.23 Ohm

Z1 angle= 84.289 deg

Voltage Fixed Control

Es = 220 kV

Ph = 0.0 deg

Frequency = 50 Hz

Transformer

The two winding 220/66 kV power transformer connects the transmission system to

the subtransmission system.


34/180

34/138

Often substation have more than one transformer operating in parallel with the bus

tie on the 220 kV closed and on the 66 kV bus open or closed according to systemrequirement, the fault level in the system and the rating of the equipment. To

simplify the study the model represents only one transformer.

Winding connection of the transformer is delta on the HV side and star on the LV

side. Tap changer operation is disabled

The transformer is represented in PSCAD according to the classical approach of

the transformer theory. Magnetising current is represented as well as core

saturation placed on the HV side. In PSCAD the HV side is the closest winding to

the core of the transformer.

Hysteresis losses in the transformer core are included. Copper losses are disabled

to reduce the number of nodes in the system. Considering that a typical X/R for a

power transformer of 100 MVA and 220/66 kV is above 30, this simplification was

considered acceptable.

TRANSFORMER DATA:

3 Phase

2 Winding Transformer

MVA Rating: 100.0 MVA

Freq= 50 Hz,

V ph-ph primary= 220 kV

Vph-ph secondary= 66kV

No Load Losses= 0.01 pu


35/180

35/138

Copper Losses=0.0 pu.

Positive Sequence Reactance= 0.15 pu.

Saturation enabled simulated on primary winding

Air core reactance= 0.3 pu

In Rush Decay Time Constant= 0.03 pu.

Knee Point Voltage: 1.25 pu

Time to release Flux Clipping=0.1s

Magnetising Current= 0.02 pu.

Overhead line

In electromagnetic simulation studies there are two main methods to representtransmission lines[21]:.

! Simplified studies generally rely on the PI Section model

! Distributed transmission line, which is most suited for transient line response

modeling using a digital computer. This method was used for this study

PSCAD operates applying the traveling wave theory but also implementing mutual

coupling between conductors and wave shape attenuation produced by the

transmission line. Transmission lines under PSCAD transient software packages

are modeled using one of the three traveling wave models:

! Bergeron Model

! Frequency Dependent Mode Model


36/180

36/138

! Frequency Dependent Phase Model

The Frequency Dependent Phase Model was selected for this study as is the most

accurate model. This model is a distributed RLC traveling wave model, which

incorporates the frequency dependence of all parameters in the model. This model

is recommended by the PSCAD Users Guide for this type of study.

The overhead line parameters are based on a typical 66 kV line as represented in

Victoria.

Conductor Data

400 mm2 aluminum

Conductor Resistance= 0.071 Ohm/km

Conductor Geometric Mean Radius 0.00982 m

Capacitor bank

The capacitor bank is implemented in Simulation 3 and 4. The capacitor bank is

connected at the end of the line to produce maximum peak value of the system

response during the transient period. The capacitance value was set under the

assumption that the load on the overhead line is 60 MVA with a power factor of

0.707.

Load Current = S/ ( !3 * V ) = 60 *106 / !3 * 66000 )= 524.8 A

P = S * PF = 52.42 MW

If PF is 0.707

Power angle = 45 degrees


37/180


38/180


39/180

39/138

Phase to ground and phase to phase to ground the fault were simulated. Current

interruption was simulated at 0 current crossing to avoid any overvoltage related tomagnetic field being trapped in the circuit. This type of overvoltage is usually

referred as current chopping[3].

Fault resistance is set to zero to evaluate the impact of the neutral earth resistor by

itself.

In summary:

Fault Type = Phase to ground and phase to phase to ground

Fault start =0.4 s to fault 0.410 s

Fault length =0.1 s

Fault ON resistance = 0.01 ohm

Fault OFF resistance = 106ohm

Current Chopping limit= 0 kA


40/180

40/138

2.5 Simulation Results

Simulation results are organized according to the six system configuration and

related simulations.

Simulation results data is attached in Appendix B.

Each simulation result Section contains two graphs. The first plot summarizes the

main objective of this investigation which is the maximum overvoltage as a function

of the reduced fault current. The fault current was calculated by applying the

symmetrical components theory as illustrated in Appendix B.

The second plot highlights the specific weight of the transitory overvoltage

component in respect to the steady state overvoltage component. Power frequency

overvoltage was subtracted from the measured overvoltage only if the recorded

maximum overvoltage took place during the fault and not at fault clearing. After

fault clearing, the power frequency overvoltage can be considered absent as the

three phases are now connected to a symmetrical voltage source.

Additional investigations were also performed on the following topics:

! Overvoltages and traveling waves

! Influence of the fault time

! Overvoltage response on healthy and faulty phases

! Frequency spectrum analysis


41/180

41/138

2.5.1 Simulation 1 : Unloaded Line - Fault at the End of the Line

TABLE 2-1SIMULATION 1 RESULTS

Line Length= 100km

NER(Ohm)

MaxMeasuredV from

PSCAD(kV)

Iph-g withNER/ Iph-g

solidlyearthed K0

V Meas /

NominalVoltage

Power

FrequencyV (pu)

Natural

Frequency V(pu)

0 126.216 1.000 1.0 2.342 1.002 2.3425 124.898 0.977 1.1 2.318 1.103 2.31810 126.607 0.946 1.3 2.349 1.204 2.34915 128.809 0.909 1.5 2.390 1.296 2.39020 128.021 0.867 1.8 2.376 1.378 2.37625 129.201 0.824 2.0 2.398 1.448 2.398

30 131.007 0.782 2.3 2.431 1.507 2.43135 131.669 0.740 2.6 2.443 1.555 2.44340 132.052 0.700 2.9 2.450 1.596 2.45045 132.457 0.663 3.2 2.458 1.629 2.45850 132.599 0.628 3.6 2.461 1.656 2.461100 103.588 0.394 6.8 1.922 1.763 1.922550 112.552 0.082 36.0 2.089 1.757 0.3321000 122.539 0.045 65.4 2.274 1.747 0.5271000000 140.632 0.000 65217.4 2.610 1.736 0.873

Line Length=10km0 90.447 1.000 1.0 1.678 1.006 1.6785 99.668 0.882 1.7 1.850 1.412 1.850

10 100.711 0.705 3.0 1.869 1.647 1.86915 99.279 0.561 4.3 1.842 1.745 1.84220 99.031 0.457 5.7 1.838 1.783 0.05425 96.724 0.383 7.1 1.795 1.797 -0.00330 96.071 0.328 8.5 1.783 1.802 -0.01935 95.465 0.286 9.8 1.772 1.802 -0.03040 95.437 0.253 11.2 1.771 1.800 -0.02945 95.512 0.227 12.6 1.772 1.797 -0.02550 95.526 0.206 14.0 1.773 1.794 -0.021100 94.901 0.105 27.9 1.761 1.771 -0.010


42/180

42/138

NER(Ohm)

Max

MeasuredV fromPSCAD(kV)

Iph-g withNER/ Iph-gsolidlyearthed K0

V Meas /NominalVoltage

PowerFrequencyV (pu)

NaturalFrequency V(pu)

550 101.180 0.019 153.0 1.878 1.740 0.1371000 107.676 0.011 278.1 1.998 1.736 0.2621000000 120.050 0.000 278041.7 2.228 1.736 0.491

Line Length=0.2km0 76.644 1.000 0.9 1.422 1.011 1.4225 86.617 0.806 2.3 1.607 1.358 1.607

10 92.223 0.564 4.4 1.711 1.455 0.25715 95.204 0.415 6.5 1.767 1.478 0.28820 96.245 0.324 8.6 1.786 1.822 -0.03625 96.587 0.264 10.8 1.792 1.817 -0.02530 96.206 0.223 12.9 1.785 1.810 -0.02535 96.501 0.192 15.1 1.791 1.804 -0.013

40 96.337 0.169 17.2 1.788 1.798 -0.01045 96.163 0.151 19.3 1.784 1.792 -0.00850 95.991 0.136 21.5 1.781 1.788 -0.007100 94.486 0.068 42.9 1.753 1.763 -0.010

550 96.337 0.012 236.1 1.788 1.738 0.0501000 102.533 0.007 429.2 1.903 1.735 0.167

1000000 118.593 0.000 429200.4 2.201 1.736 0.464

Configuration 1- Fault at the of the line- No load- Measured Overvoltage

0.0

0.5

1.0

1.5

2.0

2.5

3.0

0.00.10.20.30.40.50.60.70.80.91.0

Ifault/I ph-g solidly earthed (pu)

Overvoltage(pu)

L=100km

L = 10km

L= 0.2km

Figure 2.5: Configuration 1 Measured overvoltage result


43/180


44/180

44/138

explained using the factor K0=Z0/Z1. The implementation of the resistor will

increase Z0. As the neutral impedance has a bigger impact on the factor K0, in theshortest line the expected steady state overvoltage has a bigger increase

The results show that the maximum overvoltages recorded in the simulation with a

line length of 10km are similar to the results obtained from the simulation with a

line length of 0.2km. This characteristic is recorded throughout the study. We can

explain this characteristic by referring to the overvoltage plots as function of system

parameters of IEEE standards[18]. The different line length does not have a

particular impact on the Xco/X1. For both line lengths, the ratio is larger than 100 .

Beyond this value the expected overvoltage is considered constant.[6].

Xsource= 0.0087 pu

X Tx= 0.15 pu

X line=0.0088 Ohm/km

C= 9.4 nF/Km

Line length 10km

Xc = 1/(2f C )/ 10 = 1/( 2* 50 * 9.4 10-9)/ 10 = 33879 Ohm

X1= Xsource+ Xtx + X line = 0.0087 + 0.15 + 0.088 ( 10km) = 0.2467 pu

Z base = 484 Ohm

Xc in pu = Xc value/ Z base= 33879/484= 70 pu

Ratio Xco/X1= 70/ 0.2467= 283.7

Line length 0.2km


45/180

45/138

Xc = 1/(2f C )/0.2 = 1/( 2* 50 * 9.4 10-9)/ 0.2 = 1693996 Ohm

X1= Xsource+ Xtx + X line( 0.2km) = 0.0087 + 0.15 + 0.0018 = 0.1605 pu

Z base = 484 Ohm

Xc in pu = Xc value/ Z base= 636/484= 3500 pu

Ratio Xco/X1= 3500/ 0.1605= 21806

Line length 100 km

Xc = 1/(2f C )/0.2 = 1/( 2* 50 * 9.4 10-9)/ 100 = 3387.9 Ohm

X1= Xsource+ Xtx + X line( 100km) = 0.0087 + 0.15 + 0.88 = 0.1605 pu

Z base = 484 Ohm

Xc in pu = Xc value/ Z base= 3387.9/484= 7 pu

Ratio Xco/X1= 7/ 0.1605= 43.6

The Xco/X1 ratio for line length of 100km is also proved to be below 100.

Therefore, the line will produce a higher overvoltage.


46/180

46/138

Configuration 1- Fault at the of the line- No load- Transient Overvoltage Contribution

-0.5

0.0

0.5

1.0

1.5

2.0

2.5

3.0

0.00.10.20.30.40.50.60.70.80.91.0


Overvoltage(pu)

L=100km

L = 10km

L= 0.2km

Figure 2.6: Configuration 1 Transient overvoltage component

Figure 2.6 shows the contribution of the transient overvoltage to the maximum

recorded overvoltage.


47/180

47/138

2.5.2 Simulation 2 : Unloaded Line - Fault at the Beginning of the Line


Line Length= 100km

NER(Ohm)

MaxMeasuredV from

PSCAD(kV)

Iph-g withNER/ Iph-g

solidlyearthed K0

V Meas /

NominalVoltage

Power

FrequencyV (pu)

Natural

FrequencyV (pu)

0 72.597 1.000 0.9 1.347 1.011 1.3475 94.147 0.804 2.4 1.747 1.361 1.747

10 99.371 0.561 4.4 1.844 1.456 1.84415 101.219 0.412 6.6 1.878 1.479 1.87820 102.529 0.321 8.7 1.903 1.488 1.90325 103.542 0.262 10.9 1.921 1.817 0.10430 104.549 0.220 13.1 1.940 1.810 0.13035 106.192 0.190 15.2 1.971 1.803 0.167

40 107.291 0.167 17.4 1.991 1.797 0.194

45 108.035 0.149 19.6 2.005 1.792 0.21350 108.486 0.134 21.7 2.013 1.788 0.226100 111.955 0.068 43.4 2.078 1.763 0.315550 110.022 0.012 238.7 2.042 1.738 0.3041000 111.633 0.007 434.0 2.072 1.735 0.3361000000 111.941 0.000 433986.8 2.077 1.736 0.341

Line Length=10km0 72.781 1.000 0.9 1.351 1.011 1.3515 91.485 0.804 2.4 1.698 1.361 1.698

10 97.134 0.561 4.4 1.802 1.456 1.802

15 97.917 0.412 6.6 1.817 1.479 0.33820 97.710 0.321 8.7 1.813 1.488 0.32525 97.291 0.262 10.9 1.805 1.817 -0.01230 96.880 0.220 13.1 1.798 1.810 -0.01235 96.520 0.190 15.2 1.791 1.803 -0.012

40 96.212 0.167 17.4 1.785 1.797 -0.01245 95.950 0.149 19.6 1.781 1.792 -0.01250 95.727 0.134 21.7 1.776 1.788 -0.011100 94.570 0.068 43.4 1.755 1.763 -0.008


48/180

48/138

NER(Ohm)

Max





NaturalFrequencyV (pu)

550 108.220 0.012 238.7 2.008 1.738 0.2701000 112.946 0.007 434.0 2.096 1.735 0.3611000000 119.533 0.000 433986.8 2.218 1.736 0.482

Line Length=0.2km0 67.090 1.000 0.9 1.245 1.011 1.2455 91.473 0.804 2.4 1.697 1.361 1.697

10 97.117 0.561 4.4 1.802 1.456 0.34615 97.898 0.412 6.6 1.817 1.479 0.33820 97.681 0.321 8.7 1.813 1.488 0.32525 97.272 0.262 10.9 1.805 1.817 -0.01230 96.861 0.220 13.1 1.797 1.810 -0.01335 96.501 0.190 15.2 1.791 1.803 -0.01340 96.193 0.167 17.4 1.785 1.797 -0.01245 95.931 0.149 19.6 1.780 1.792 -0.01250 95.708 0.134 21.7 1.776 1.788 -0.012100 94.570 0.068 43.4 1.755 1.763 -0.008550 108.314 0.012 238.7 2.010 1.738 0.272

1000 112.946 0.007 434.0 2.096 1.735 0.3611000000 113.404 0.000 433986.8 2.104 1.736 0.368

Configuration 2- Fault at the bus- No load- Measured Ov ervoltage

0.0

0.5

1.0

1.5

2.0

2.5

0.00.10.20.30.40.50.60.70.80.91.0


Overvoltage(p

u)

L=100km

L = 10kmL= 0.2km

Figure 2.7- Configuration 2 Measured overvoltage result


49/180


50/180

50/138

2.5.3 Simulation 3 : Capacitor Bank. Fault at the End of the Line


Line Length= 100km

NER(Ohm)

MaxMeasuredV fromPSCAD(kV)





0 187.403 1.000 1.0 3.478 1.002 2.4765 185.285 0.977 1.1 3.438 1.102 2.33610 184.351 0.946 1.3 3.421 1.204 2.21715 184.572 0.909 1.5 3.425 1.296 2.12920 185.656 0.867 1.8 3.445 1.378 2.06725 187.384 0.824 2.0 3.477 1.448 2.02930 189.670 0.782 2.3 3.520 1.507 2.01335 192.279 0.740 2.6 3.568 1.555 2.01340 195.097 0.700 2.9 3.620 1.596 2.025

45 199.190 0.663 3.2 3.696 1.629 2.06850 203.527 0.628 3.6 3.777 1.656 2.121100 245.786 0.394 6.8 4.561 1.763 2.798550 406.039 0.082 36.0 7.535 1.757 5.7781000 453.542 0.045 65.4 8.416 1.747 6.6691000000 526.382 0.000 65217.4 9.768 1.732 8.036

Line Length=10km0 110.323 1.000 1.0 2.047 1.006 1.0415 123.489 0.882 1.7 2.292 1.412 0.87910 127.826 0.705 3.0 2.372 1.647 0.72515 128.511 0.561 4.3 2.385 1.745 0.64020 124.144 0.457 5.7 2.304 1.783 0.52125 154.689 0.383 7.1 2.871 1.797 1.07330 160.269 0.328 8.5 2.974 1.802 1.17235 164.680 0.286 9.8 3.056 1.802 1.25440 168.268 0.253 11.2 3.123 1.800 1.32345 171.273 0.227 12.6 3.178 1.797 1.38150 173.731 0.206 14.0 3.224 1.794 1.430100 186.406 0.105 27.9 3.459 1.771 1.688


51/180

51/138

NER(Ohm)

Max






550 198.797 0.019 153.0 3.689 1.740 1.9491000 200.148 0.011 278.1 3.714 1.736 1.9781000000 182.224 0.000 278041.7 3.381 1.736 1.645

Line Length=0.2km0 109.095 1.000 0.9 2.024 1.011 1.0145 135.208 0.806 2.3 2.509 1.358 1.151

10 135.715 0.564 4.4 2.518 1.455 1.06415 127.208 0.415 6.5 2.361 1.478 0.88220 148.710 0.324 8.6 2.760 1.822 0.93725 152.521 0.264 10.8 2.830 1.817 1.01330 162.300 0.223 12.9 3.012 1.810 1.20235 166.755 0.192 15.1 3.094 1.804 1.29140 170.288 0.169 17.2 3.160 1.798 1.36245 173.169 0.151 19.3 3.213 1.792 1.42150 175.588 0.136 21.5 3.258 1.788 1.471100 188.724 0.068 42.9 3.502 1.763 1.739550 200.208 0.012 236.1 3.715 1.738 1.977

1000 201.435 0.007 429.2 3.738 1.735 2.0031000000 203.044 0.000 429200.4 3.768 1.736 2.031

Configuration 3- Fault at the end of the line- Cap Bank connected at the end of the line- Measured

Overvoltage

0.0

2.0

4.0

6.0

8.0

10.0

12.0

0.00.10.20.30.40.50.60.70.80.91.0


Overvoltage(pu)

L=100km

L = 10km

L= 0.2km



52/180

52/138

Configuration 3- Fault at the end of the line- Cap Bank connected at the end of the line-

Transient Overvoltage Contribution

0.0

1.0

2.0

3.0

4.0

5.0

6.0

7.0

8.0

9.0

0.00.10.20.30.40.50.60.70.80.91.0Ifault/I ph-g so lidly earthed (pu)

Overvoltage(pu)

L=100km

L = 10km

L= 0.2km


The dangerous effects of capacitor switching are well known in the engineering

field [23]. The results obtained with this simulation show that, following a fault in the

network with no or small loads and a capacitor bank connected to the grid

represent a danger to the insulation of the power system.

The overvoltages measured in this simulation are the highest of the entire study

reported here. The results are in line with previous studies [5, 14] except for

results obtained with line length of 100 km . This is due to the combination of the

following system parameters:

!

Length of the line

! NER size

These parameters affect the3

00 XC

R " transient overvoltages condition. If this

condition is not met, dangerous fault restrikes are generated [9]. Is it worth noting

that field experience of arcing ground fault above 7 pu are quite rare as the

insulation of the power system usually breaks down between 6 to 7 pu [9]


53/180

53/138

Line length impact is shown in Figure 2.9 and 2.10. Instead, to review the impact of

the NER size further simulations were performed. Four snapshots with a NER of10, 100, 1000 and 1000000 Ohm were taken as shown in Figures 2.11 to 2.14.

-80

-60

-40

-20

0

20

40

60

80

100

Voltage

(kV)

Vs

Figure 2.11: Configuration 3 Transient overvoltage Snapshot with NER of 10 Ohms

-300

-200

-100

0

100

200

300

VV


54/180

54/138

Figure 2.14: Configuration 3 Transient overvoltage Snapshot with NER of 1000000 Ohms

The graphs show that the increase of the resistor will increase the transient

overvoltages.

The dangerous transient overvoltages are also due to the conservative approach in

selecting the size of the capacitor bank. A further simulation is conducted to

validate this theory. A smaller capacitor bank of 15 MVA was used for the analysis.

Configuration 3- Line Length 100km- Comparison between 30MVAR and 15MVAR Capacitor BankResults- Measured Overvoltage

0.0

2.0

4.0

6.0

8.0

10.0

12.0

0.00.10.20.30.40.50.60.70.80.91.0


Overvoltage(pu)

15MVAR

30MVAR

Figure 2.15: Configuration 3 30 and 15 MVAR Cap bank overvoltage comparison


55/180

55/138

The system response with a smaller capacitor gives a result which is acceptable

and in line with previous studies [5, 19]. Maximum overvoltage with the insulatedsystem is 4.1 pu.


56/180

56/138

2.5.4 Simulation 4 : Capacitor Bank. Fault at the Beginning of the Line


Line Length= 100km

NER

(Ohm)

MaxMeasuredV fromPSCAD

(kV)

Iph-g withNER/ Iph-gsolidly

earthed K0

V Meas /Nominal

Voltage

PowerFrequency

V (pu)

NaturalFrequency

V (pu)0 73.660 1.000 0.9 1.367 1.011 1.3675 102.229 0.804 2.4 1.897 1.361 1.89710 110.677 0.561 4.4 2.054 1.456 2.05415 120.455 0.412 6.6 2.235 1.479 2.23520 139.620 0.321 8.7 2.591 1.488 2.59125 156.662 0.262 10.9 2.907 1.817 2.90730 171.647 0.220 13.1 3.185 1.810 3.18535 185.697 0.190 15.2 3.446 1.803 3.44640 198.103 0.167 17.4 3.676 1.797 3.67645 209.035 0.149 19.6 3.879 1.792 3.879

50 219.140 0.134 21.7 4.067 1.788 4.067100 282.383 0.068 43.4 5.240 1.763 5.240550 373.638 0.012 238.7 6.934 1.738 6.9341000 431.569 0.007 434.0 8.009 1.735 8.0091000000 553.568 0.000 433986.8 10.272 1.736 10.272

Line Length=10km0 73.885 1.000 0.9 1.371 1.011 0.3605 104.589 0.804 2.4 1.941 1.361 0.58010 115.953 0.561 4.4 2.152 1.456 0.69615 126.137 0.412 6.6 2.341 1.479 0.86220 129.761 0.321 8.7 2.408 1.488 0.92025 140.232 0.262 10.9 2.602 1.817 0.78530 143.196 0.220 13.1 2.657 1.810 0.84735 147.174 0.190 15.2 2.731 1.803 0.92840 151.560 0.167 17.4 2.812 1.797 1.01545 155.590 0.149 19.6 2.887 1.792 1.09550 158.744 0.134 21.7 2.946 1.788 1.158100 173.853 0.068 43.4 3.226 1.763 1.463550 185.633 0.012 238.7 3.445 1.738 1.707


57/180


58/180


59/180

59/138

2.5.5 Simulation 5 : Light inductive Load. Fault at the End of the Line.


Line Length= 100km

NER(Ohm)

MaxMeasured

V fromPSCAD(kV)

Iph-g with

NER/ Iph-gsolidlyearthed K0



NaturalFrequency V (pu)

0 123.996 1.000 1.0 2.301 1.002 2.3015 121.497 0.977 1.1 2.255 1.102 2.25510 127.242 0.946 1.3 2.361 1.204 2.36115 124.414 0.909 1.5 2.309 1.296 2.30920 125.584 0.867 1.8 2.330 1.378 2.33025 126.472 0.824 2.0 2.347 1.448 2.34730 127.242 0.782 2.3 2.361 1.507 2.36135 127.753 0.740 2.6 2.371 1.555 2.371

40 128.210 0.700 2.9 2.379 1.596 2.37945 128.382 0.663 3.2 2.382 1.629 2.38250 128.563 0.628 3.6 2.386 1.656 2.386100 125.448 0.394 6.8 2.328 1.763 2.328550 106.256 0.082 36.0 1.972 1.757 0.2151000 114.514 0.045 65.4 2.125 1.747 0.378

1000000 131.252 0.000 65217.4 2.436 1.732 0.704

Line Length=10km0 89.031 1.000 1.0 1.652 1.006 1.6525 98.999 0.882 1.7 1.837 1.412 1.837

10 99.542 0.705 3.0 1.847 1.647 1.84715 98.172 0.561 4.3 1.822 1.745 1.82220 97.019 0.457 5.7 1.800 1.783 0.01725 95.755 0.383 7.1 1.777 1.797 -0.02130 95.531 0.328 8.5 1.773 1.802 -0.02935 95.024 0.286 9.8 1.763 1.802 -0.03840 94.367 0.253 11.2 1.751 1.800 -0.04945 94.358 0.227 12.6 1.751 1.797 -0.04650 94.180 0.206 14.0 1.748 1.794 -0.046100 93.855 0.105 27.9 1.742 1.771 -0.029


60/180

60/138

550 99.236 0.019 153.0 1.841 1.740 0.101

NER(Ohm)






1000 109.857 0.011 278.1 2.039 1.736 0.3021000000 241.638 0.000 278041.7 4.484 1.736 2.748

Line Length=0.2km0 74.717 1.000 0.9 1.387 1.011 1.3875 85.939 0.806 2.3 1.595 1.358 1.595

10 91.418 0.564 4.4 1.696 1.455 0.24215 94.402 0.415 6.5 1.752 1.478 0.27320 95.404 0.324 8.6 1.770 1.822 -0.05225 95.743 0.264 10.8 1.777 1.817 -0.04030 95.531 0.223 12.9 1.773 1.810 -0.03735 95.640 0.192 15.1 1.775 1.804 -0.02940 95.477 0.169 17.2 1.772 1.798 -0.02645 95.301 0.151 19.3 1.768 1.792 -0.02450 95.131 0.136 21.5 1.765 1.788 -0.022100 94.017 0.068 42.9 1.745 1.763 -0.019550 97.021 0.012 236.1 1.800 1.738 0.062

1000 107.684 0.007 429.2 1.998 1.735 0.2631000000 126.718 0.000 429200.4 2.351 1.736 0.615

Configuration 5- Fault at the of end the line- Inductive Load connected at the end of the

line- Measured Overvoltage

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

5.0

0.00.10.20.30.40.50.60.70.80.91.0


Overvoltage(pu) L=100km

L = 10km

L= 0.2km



61/180

61/138

Configuration 5- Fault at the end of the line- Inductive Load connected at the end of the line-


-0.5

0.0

0.5

1.0

1.5

2.0

2.5

3.0

0.00.10.20.30.40.50.60.70.80.91.0


Overvoltage(pu)

L=100km

L = 10km

L= 0.2km


Maximum overvoltages recorded are similar to the results from Simulation 1. It is

likely that this inductive load was not large enough to produce any relevant impact

on the overvoltages. Therefore, not further analysis was conducted using the

results of this system configuration.


62/180

62/138

2.5.6 Simulation 6 : Energized 66/22kV Transformer Fault at the 66kV

Bus.


NER(Ohm)






0 54.929 1.000 0.9 1.017 1.011 0.0065 91.224 0.804 2.4 1.692 1.361 0.33110 96.846 0.561 4.4 1.792 1.456 0.33615 97.661 0.412 6.6 1.812 1.479 0.33320 97.464 0.321 8.7 1.806 1.488 0.31825 97.053 0.262 10.9 1.800 1.817 -0.01730 96.646 0.220 13.1 1.793 1.810 -0.01735 96.287 0.190 15.2 1.788 1.804 -0.01640 95.982 0.167 17.4 1.786 1.798 -0.01245 95.721 0.149 19.6 1.776 1.792 -0.01650 95.499 0.134 21.7 1.772 1.788 -0.016

100 94.343 0.068 43.4 1.763 1.763 0.000550 93.211 0.012 238.7 1.729 1.738 -0.0091000 93.089 0.007 434.0 1.727 1.735 -0.0081000000 92.838 0.000 433986.8 1.723 1.732 -0.009

Configuration 6- Fault at the bus- Tx 30 MVA delta/star at the bus- Measured Overvoltage

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

2.0

0.00.10.20.30.40.50.60.70.80.91.0


Overvoltage(pu)



63/180

63/138

Configuration 6- Fault at the bus- Tx 30 MVA at the bus-


-0.1

0.0

0.1

0.1

0.2

0.2

0.3

0.3

0.4

0.4

0.00.10.20.30.40.50.60.70.80.91.0

Ifault/I ph-g solidly earthe d (pu)

Overvoltage(pu)

Figure 2.21: Configuration 6- Transient overvoltage result

Overvoltages recorded in this study are relatively lower than in the other

simulations. Maximum overvoltage reaches a maximum of 1.8pu as phase to

ground fault decreases below 0.6 pu.

Please note that the transient overvoltage component disappears as the phase to

ground fault decreases below 0. 25 pu. Therefore, for high resistor values the

overvoltages are only produced by the power frequency voltage.

No further analysis was conducted using the data from this simulation

2.5.7 Overvoltages and Traveling Waves

This section documents the propagation of the voltage traveling waves along the

transmission line. A series of snapshots are presented to describe the behavior of

the overvoltages using the data from Simulation 1 No load at the end of the line-

and with the line length of 100 km


64/180

64/138

Bus Current,Bus Voltage : Graphs

Time ... 0.3850 0.3900 0.3950 0.4000 0.4050 0.4100 0.4150 0.4200 0.4250 0.4300

-1.50

-1.00

-0.50

0.00

0.50

1.00

1.50

Current(kA)

Is

-100

-80

-60

-40

-20

0

20

40

60

80

Voltage

(kV)

Vs

Figure 2.22: Configuration 1 - Maximum Overvoltage - Phase to Ground Fault NER 0 Ohm

Snapshots were recorded at fault inception and fault clearing time. The graph also

shows that the transient overvoltage at higher frequency is superimposed on a

lower frequency overvoltage. The overvoltage has a period of 0.7 ms which is

equivalent to a frequency of 1400 Hz. The high frequency component decays inapproximately 10ms.

Traveling waves are triggered at the same time on both healthy phases and travel

on the conductors. The steps pattern of the transient overvoltage is the effect of the

traveling waves that reflect back and forward between the two ends of the line.


65/180

65/138


Time ... 0.5050 0.5075 0.5100 0.5125 0.5150 0.5175 0.5200 0.5225 0.5250 0.5275

-1.50

-1.00

-0.50

0.00

0.50

1.001.50

Current(kA)

Is

-100

-75

-50

-25

0

25

50

75

100

125

Voltage

(kV)

Vs

Figure 2.23: Configuration 1- Snapshot of after clearing overvoltage- Phase to Ground Fault NER

0 Ohm

Transient overvoltage superimposed on the power frequency voltage has a

frequency of 590 Hz. Maximum overvoltage appears on the phase which has the

highest voltage at the clearing time. The overvoltage is produced by the energy

stored in the electric field of the capacitance that will discharge at fault clearing

time.


66/180

66/138


Time ... 0.490 0.500 0.510 0.520 0.530 0.540 0.550

-1.50

-1.00

-0.50

0.00

0.50

1.001.50

Current(kA)

Is

-100

-50

0

50

100

150

Voltage

(kV)

Vs

Figure 2.24: Configuration 1- Snapshot of Maximum Overvoltage- Phase to Phase to Ground Fault

NER 0 Ohm

In the phase to phase ground fault the maximum overvoltage is recorded at the

fault clearing time. Frequency of the superimposed transient overvoltage is 590 Hz

as in the phase to ground faults. The overvoltages on each phase have the same

frequency. This transient overvoltage decays after 30ms.


67/180


68/180

68/138

As the NER approaches 50 Ohm, the pattern of the plots remains similar to plots

with NER equal to 0 Ohm. Overvoltage frequencies are 590 Hz in all three phases.


Time ... 0.4000 0.4050 0.4100 0.4150 0.4200 0.4250

-0.100

-0.050

0.000

0.050

0.100

0.150

0.200

Current(kA)

Is

-125

-100

-75

-50

-25

0

25

50

75

100

Voltage

(kV)

Vs

Figure 2.27: Configuration 1- Snapshot of Maximum Overvoltage- Phase to Ground Fault NER

1000 Ohm

The large size of the resistor now makes the phase to ground fault as the most

prominent overvoltage. Traveling waves are visible in the current and voltage plots.

2.5.8 Impact of Time of Fault with the Overvoltage

This investigation highlights the relation between the time of fault and the

amplitude of the measured overvoltage. Plots are divided into phase to ground fault

and phase to phase to ground fault.


69/180

69/138

Simulation 1- Measured Overvoltage

Ph-g fault- All Data Plot

0

0.5

1

1.5

2

2.5

3

0 50 100 150 200 250 300 350

Sample

Overvoltag L=100km

L=10km

L=0.2km

Solidly Earthed

SystemInsulated

System

Figure 2.28: Configuration 1- All Data Plot


Ph-g Fault- All Data Plot

0

0.5

1

1.5

2

2.5

0 50 100 150 200 250 300 350

Sample

Overvoltag L=100km

L=10km

L=0.2km

Solidly Earthed

SystemInsulated

System

Figure 2.29: Configuration 2 All Data Plot



0

2

4

6

8

10

12

0 50 100 150 200 250 300 350

Sample

Overvoltag L=100km

L=10km

L=0.2km

Solidly Earthed

SystemInsulated

System

Figure 2.30: Configuration 3 All data plot


70/180

70/138



0

2

4

6

8

10

12

0 50 100 150 200 250 300 350

Sample

Overvoltag L=100km

L=10km

L=0.2km

Solidly Earthed

SystemInsulated

System

Figure 2.31: Configuration 4- All data plot

Simulation 1- Measured OvervoltagePh-Ph-G Fault- All Data Plot

0

0.5

1

1.5

2

2.5

3

0 50 100 150 200 250 300 350

Simulation

Overvoltage(pu)

Series1

Series2

Series3Solidly EarthedSystem

InsulatedSystem



0

0.5

1

1.5

2

2.5

0 50 100 150 200 250 300 350

Simulation

Overvoltage(pu)

L=100km

L=10km

L=0.2kmSolidly Earthed

System

InsulatedSystem



71/180

71/138


0

0.5

1

1.5

2

2.5

3

3.5

4

4.5

5

0 50 100 150 200 250 300 350

Simulation

Overvoltage(pu)

L=100km

L=10km

L=0.2kmSolidly E arthedSystem

InsulatedSystem



Ph-Ph-G Fault- All Data Plot

0

1

2

3

4

5

6

7

8

0 50 100 150 200 250 300 350

Simulation

Overvoltage(pu)

L=100km

L=10km

L=0.2kmSolidly EarthedSystem

InsulatedSystem


These graphs are produced by using the maximum overvoltage recorded on each

simulation. It is reminded that 11 simulations were performed for each NER sizeconfiguration. Besides, 1 ms additional delay time was applied on each simulation.

Points were plotted sequentially, starting on the simulation with a NER of 0 Ohm to

the maximum value of the NER, which corresponds to the insulated system.

In these eight plots, Figure 2.28 to Figure 2.35, it is possible to recognize the 15

peaks related to the 15 resistor sizes used in the simulations.


72/180

72/138

Phase to ground faults for transmission line of 10 and 0.1km showed a smoother

pattern with lower peak values. Only in Simulation 2 there is an irregularity in themaximum overvoltage. Overall, the maximum overvoltage in the phase to ground

fault is therefore less dependent on the time of fault.

Phase to ground fault maximum overvoltages show a similar pattern across the

three different subtransmission line lengths. Maximum overvoltages appear only for

a brief period of time and at the same time in all the three line lengths.

The 8 plots also show similar results for simulations conducted with

subtransmission line of 10km and 0.1km.

These graphs confirm that time of fault is relevant in determining the maximum

overvoltage during phase to phase to ground faults.

2.5.9 Overvoltage Response on Phase a ,b and c

The section investigates the relationship between faults and different overvoltage

responses on each phase.

This study was conducted only for subtransmission line of 100km in the simulation

1, 2, 3 and 4. Graphs are divided in phase to ground faults and phase to phase to

ground faults.

On PSCAD software package phase to ground fault takes place on phase a and

phase to phase to ground fault takes place on phases a and b.


73/180

73/138

Simulation 1-Measured Overvoltage

Phase Comparison-Phase to Ground Fault- L=100km

0

0.5

1

1.5

2

2.5

3

0 50 100 150 200 250 300 350

Simulation

Overvoltage(pu)

Phase a

Phase b

Phase c

Solidly Earthed

System

Insulated

System

Figure 2.36: Configuration 1 Phase comparison



0

0.5

1

1.5

2

2.5

0 50 100 150 200 250 300 350

Simulation

Overvoltage(p)

Phase a

Phase b

Phase c

Solidly Earthed

System

Insulated

System




0

2

4

6

8

10

12

0 50 100 150 200 250 300 350

Simulation

Overvoltage(pu)

Phase a

Phase b

Phase c

Solidly Earthed

System

Insulated

System



74/180

74/138

Simulation 4-Measured OvervoltagePhase Comparison-Phase to Ground Fault- L=100km

0

2

4

6

8

10

12

0 50 100 150 200 250 300 350

Simulation

Overvoltage(pu)

Phase a

Phase b

Phase c

Solidly EarthedSystem

InsulatedSystem



Phase Comparison-Phase to Phase to Ground Fault- L=100km

0

0.5

1

1.5

2

2.5

3

0 50 100 150 200 250 300 350

Simulation

Overvoltage(pu)

Phase a

Phase b

Phase c

Solidly Earthed

System

InsulatedSystem



Phase Comparison-Phase to Phase to Ground Fault- L=100km

0

0.5

1

1.5

2

2.5

0 50 100 150 200 250 300 350

Simulation

Overvoltage(pu)

Phase a

Phase b

Phase c

Solidly Earthed

System

Insulated

System



75/180


76/180

76/138

As noted in Section 2.4.9, the plots for the phase to phase to ground faults show

that the time of fault is a key element to determine in which phase the maximumovervoltage is recorded.

2.5.10 Overvoltage Harmonic Response

A further analysis was conducted on the frequency spectrum of the transient

overvoltage. The study was done using the Fast Fourier Transform (FFT) function,

which determine the harmonic magnitude and phase of the input signal as a

function of time[21].

Base frequency was set to 12.5 Hz to detect noise elements lower than power

frequency voltage.

Here below few snapshots of the healthy and faulty phases for a phase to ground

fault following a fault are provided. Snapshots were captured after 10 ms fault

injection.

Referring to the graphs left hand side spectrum refers to 12.5 Hz component and

right hand side spectrum refers to 322.5 HZ noise element. The number in the plot

frame refers to the voltage magnitude in the first frequency band, 12.5 Hz.

Ea60.0

0.0

[1] 1.79434

Figure 2.44: Configuration 1 Harmonic analysis NER 0 ohmsafter 10ms


77/180

77/138

Ea60.0

0.0

[1] 3.50508


Ea60.0

0.0

[1] 6.13753


From the graphs we can observe that voltage noise appears at frequencies higher

and lower values in respect of the power frequency voltage The major components

are in the band close to 50 Hz. Components with lower magnitude, less than 3 %

of nominal voltage, are visible on high frequency up to 300 Hz.

The harmonics were recorded on the healthy phase. The graphs show that thenoise is not limited to odd harmonics. Therefore, filtering this natural frequency

component could be problematic and slow down the information process of the

digital relays.


78/180

78/138

2.6 Conclusion

1. Results obtained from this study are in line with the existing literature and

previous studies in this field. Maximum overvoltages are below 2.5 pu for a

ground fault in an unloaded subtransmission line and below 4 pu in an

unloaded line with a capacitor bank connected at the end of the line

2. Maximum overvoltage is influenced by the power system parameters. The

size of the neutral earth resistor itself does not give a clear indication of the

system response and expected maximum overvoltage during a ground fault

3. Increasing the size of the neutral earth resistor controls the maximum

overvoltage on unloaded line as the resistor has a positive impact in limiting

the natural frequency overvoltage until the fault current is equal to 10 % of

the fault current with a solidly earthed system

4. The limitation of the transitory overvoltage component is strictly connected

to the increase of the ratio R0/X0 until the R0


79/180

79/138

analysis and related transient studies should be carefully considered for

long subtransmission lines as is usually done for transmission lines

8. There is also a consistent similarity of results for line length of 10km and

0.2km. This characteristic should be taken into consideration to optimize the

work during planning and design activities for short and medium length

subtransmission lines

9. Transient overvoltages appear in the faulty and healthy phases. There is a

relation between time of fault and location where the maximum overvoltages

takes place

10. There is not a strong relation between the time of fault and the magnitude of

the maximum overvoltage during phase to ground faults. Relation between

the magnitude of the maximum overvoltage and the time of fault is only

relevant on the phase to phase to ground faults

11. Considering the overvoltage component at higher frequency, this study

shows that beyond the debate of effective or non effective grounding, design

of the neutral earth resistors could safely aim to limit the ground fault current

to 0.1pu of the fault current in a solidly earthed system without

compromising the insulation level of the system


80/180

80/138

3 CVT and Transient Overvoltages

3.1 Overview

Capacitive voltage transformer, CVT, represents the most common method of

voltage source for the impedance protection relays installed for transmission and

subtransmission power systems.

Until few years ago the use of the CVT was limited to extra high voltage system. In

recent years the use of the CVT has become more popular even for applicationwith nominal voltage down to 33kV (as applied by Energex and Ergon in

Queensland) due the lower capital cost compared to the traditional voltage

transformer, VT. CVTs also offer the advantage to couple power line carrier to

inject teleprotection signals between substations.

During a fault, ideally, the low voltage source provided by the CVT should be an

exact copy of the primary voltage. Unfortunately this is not the reality. CVT

contains a large number of stored energy components which must charge and

discharge during voltage changes [24-29]. The transfer of energy between

capacitances and inductances within the CVT requires some time to take place.

Therefore, the CVT introduces some transients that affect the magnitude and

shape of the voltage signal.

CVT creates some problems to the distance relays especially during faults that

severely depress the voltage at the relay location. Faults located at the beginning

of the protected line can bring the voltage down to few percents of the nominal

voltage and this could affect accuracy and operating time of the relay.

This section aims to investigate the CVT transient response during phase to

ground fault and the impact that this disturbance can have on the distance relay.


81/180


82/180

82/138

The tuning reactor is required to cancel out the value of the capacitive divider at

the system frequency [24] . Therefore, the reactor prevents the phase shiftbetween the voltage signal at the transmission line and the voltage fed to the relay.

The size of the reactor is tuned for the system frequency of 50 Hz. In this study the

reactor is 111 H.

Voltage transformer

The voltage transformer reduces the voltage from 10-15 kV to a voltage suitable for

the relay which is usually 100 or 110 V. Copper and iron losses were taken into

consideration by adding in the CVT model a resistance of 2000 Ohm.

Ferroresonance Suppression Circuit

The ferroresonance suppression circuit is installed in the low voltage side of the

voltage transformer to prevent dangerous overvoltages due to the saturation of the

iron core of the step down transformer.

Two types of ferroresonance suppression circuits are normally used by CVT

manufacturers [24, 25, 27]:

! Active ferroresonance suppression circuit

! Passive ferroresonance suppression circuit

The active ferroresonance suppression circuit consists of a parallel tuning circuit in

series with a damping resistor. The LC parallel circuit behaves as an open circuit at

the nominal power frequency to prevent shunt current flowing through the circuit.

During transient condition the tuning circuit will decrease the impedance to the

value of the damping resistor. The resistor will attenuate the energy of the transient

surge [24]


83/180

83/138

The passive ferroresonance suppression circuit has a resistor, a saturable inductor

and air gap loading resistor. In normal condition the voltage is not high enough toflash the air gap. During transient conditions, the air gap will flash over through the

resistor and will attenuate the transient energy.

Existing literature and previous study demonstrated that the passive

ferroresonance circuit has better transient performance than the active device [24,

25, 27, 28, 31, 32]. However, as the study aims to test the CVT under the worst

case scenario, the active suppression circuit was implemented.

Device Data

R= 77379 Ohms

C=29 nF

L=111 H

CVT Burden

The burden connected at the end of the circuit represents the static or numerical

digital relays. Modern relays have a small burden. Typically the burden is below 5

VA .

The value of the burden is 1000 Ohms


84/180

84/138

3.3 Simulation Settings

Fault time

System faults can happen at any time. Considering the two voltage magnitude end

range, the fault could take place at 0 voltage or at maximum voltage. Fault at zero

voltage are less common than faults at the maximum voltage. In fact, this type of

fault could only happen during lightning or close of the circuit breaker on fault

condition. Faults at the maximum voltage usually take place when insulation failure

occurs and are statistically more likely to happen.

Simulations were conducted with fault at zero crossing of the primary voltage. At

this time the energy store in the capacitor is at the maximum and this condition will

generate the worst voltage transient conditions. The sudden change in the system

will release the energy stored in the active elements of the CVT creating the worst

transient overvoltage condition. These voltage oscillations are composed of lowerand higher frequency components.

Network and Source Settings

66 kV subtransmission line model is as per the model used in Section 1

(Overvoltage). The line model was set to 2 km and 10 km. Short lines were

selected to increase the source impedance ratio, SIR, and testing the performance

of the CVT with a higher SIR. It is reminded that SIR is the ration between the

source impedance and the line impedance at the relay point.


85/180

85/138

The source impedance was set to determine a starting SIR of 1 and 10 without the

implementation of the NER. Then, the neutral earth resistor was increased toobtain the desired SIR.

To conclude, four network models were investigated:

! 10km line length- SIR of 1( No NER), 5, 10, 20, 30 and 60

! 10km line length- SIR of 10( No NER), 20, 30 and 60

! 2 km line length- SIR of 1( No NER), 5, 10, 20, 30 and 60

! 10km line length- SIR of 10( No NER), 20, 30 and 60


86/180

86/138

3.4 Simulation Results

Simulations results are provided in Appendix D and summarised in the graphs here

below:

Transient Overvoltage- 10km line- Initial SIR 1

0.00

5.00

10.00

15.00

20.0025.00

30.00

35.00

40.00

45.00

50.00

10 30 50 70

Time (ms)

Overvoltage(pu) SIR 1

SIR 5

SIR 10

SIR 20

SIR 30

SIR 60

Figure 3.1: Peak overvoltage after fault- Initial SIR 1- Line length 10 km

Transient Overvoltage- 10km line- Initial SIR 10

0.00

5.00

10.00

15.00

20.00

25.00

30.00

35.00

40.00

45.00

50.00

10 30 50 70

Time (ms)

Overvoltage(pu)

SIR 10

SIR 20

SIR 30

SIR 60

Figure 3.2: Peak overvoltage after fault- Initial SIR 10- Line length 10 km


87/180

torelli za ucinke

Documents