CHAPTER 2

COMPREHNSIVE REVIEW OF DIFFERENT BUSBAR PROTECTION

SCHEMES

2.1 INTRODUCTION
Modern electrical power system is spreaded over large geographical area with great
complexity. Power quality requirements resulting from such great complex electrical
network have motivated the improvement of fault location methods and development of
protection scheme for busbar. Busbar is a vital, often overlooked, part of the power
system. Even though faults on busbar are very rare, they can lead to disturbances & major
shutdown of power system network. Thus, protection of busbar requires highly reliable,
fast and high speed relaying scheme. With the fast changing technologies in the field of
busbar protection, new protection concepts addressing new technologies are coming to the
market. Therefore, there is an urgent need to keep track of different past and present
relaying techniques, international standards, design and service experiences of
digital/numerical relays and recent algorithmic developments taking place in the field of
busbar protection.
This chapter gives a bibliographical survey, comprehensive review of different
busbar relaying schemes and conceptual aspect of research and developments in the field
of busbar protection during the past 50 years, based on numerous published articles.
Different busbar protection techniques have been studied and a comparative evaluation of
them is undertaken with their relative advantages and disadvantages for identification of
optimum technique as applied to specific bus configuration.
2.2 BUSBAR FAULTS AND PROTECTION REQUIREMENT
Protection engineer try to implement a dedicated bus zone protection scheme from
the various schemes developed by researchers. If it is not implemented perfectly, the
clearing of bus fault will be performed by the backup protection provided to the lines
terminating at the bus. The high fault levels associated with a busbar require fast
protection. Typical fault clearing time should be less than 100 ms; with fast breakers, this
means that the measuring time should be about 20 to 30ms.
The prime requirement of the busbar protection scheme is the identification of fault
in a particular region and disconnection of circuit breakers associated with that particular
region so that healthy section of power system network remains unaffected. This
17
discriminating feature of the busbar protection scheme minimizes interruption of the plant.
Hence, it is justified that discrimination on the basis of time graded relays is not a good
choice, and hence, there is a need to develop genuine protection scheme for the busbar.
Busbar protection scheme should not operate for any external (through) fault, as
otherwise it unnecessarily trips other healthy lines connected to the bus. This is a very
important feature of the busbar protection scheme with respect to the stability of the power
system.
It is to be noted that all these requirements cannot be achieved completely by any of
the schemes without affecting the other requirements. Therefore, there is always a
compromise between many factors such as speed versus selectivity and stability versus
dependability.
Whenever bus fault occurs, it results in severe disturbances as clearance of this fault
requires tripping of all the breakers of the lines connected to the faulted bus. Moreover, the
danger caused because of unprotected bus during bus fault is found to be very severe
[110], [206]. Thus, in order to prevent hazard to the system, busbar protection scheme is
designed and implemented in such a way that unwanted tripping should not occur. By
involving bus sectionalizer and different bus arrangement, the number of circuits that must
be opened for a bus fault is minimized and hence, eliminating total interruption of power
[91]. Bus faults have been observed to be relatively rare around 8% of all faults compared
with line faults which are over 60% [195], [10], [183]. It has been observed from the
widely published literature that most of the bus faults are ground faults having 67% are
single line to ground faults, 15% are double line to ground faults and 19% are triple line to
ground faults [62].
Busbar faults can be broadly classified as:
(i) Fault due to insulation failure caused from damaged CBs or insulators
(ii) Arcing and insulator flashover caused by over voltages
(iii) Faulty handling of switching equipments especially earthing switch
(iv) Dropping off of metal parts across busbars
Thus, the protection provided must possess the following characteristics
(i) It is fast enough to minimize damage and maintain system stability
(ii) It provides correct operation during an internal fault
(iii) It maintains stability by avoiding maloperation in case of an external fault
(iv) It supervises all CTs, which provides actuating quantity to the relay in case of fault

18
2.3 CLASSIFICATION OF BUSBAR PROTECTION SCHEMES

The consolidated review of different techniques used by utilities along with their
relative merits and shortcomings are presented here. These techniques are classified in
three broad categories namely

(i) Non-unit protection

(ii) Unit protection and
(iii) Commercial techniques

Different techniques have been used by the utilities in order to detect faults on
busbar in power system. The schemes categorized under unit protection are:

(i) Directional comparison technique

(ii) Differential protection technique and
(iii) Techniques based on Artificial Intelligence (AI)

Further, techniques used in commercial relay are sub-classified into five groups:

(i) Microprocessor based scheme

(ii) Digital differential scheme
(iii) Nneural network based scheme
(iv) Traveling-wave based scheme
(v) Wavelet transform based scheme

2.3.1 Non-Unit Protection Scheme

The non-unit protection scheme is governed by back-up over current, earth fault and
distance relays. The high fault levels associated with busbars require that protection must
be very fast. Typical fault clearing time should be less than 100 ms and with fast breakers
it should be about 15 to 30 ms [38]. In order to minimize the interruption of the plant the
protection system must correctly identify the area of the fault and open only the necessary
and minimum number of breakers. However, because of speed requirements, these relays
are usually found to be non-discriminative and slow in operation. Further, the vagrancy of
the scheme applied to busbar protection can cause unnecessary tripping when relay
spuriously trips thereby causing discontinuity of supply. In addition, it can cause havoc to
equipment if back-up relays do not operate quickly. Hence, the philosophy of back-up

19
protection to clear busbar faults can be confined to radial system only. It is therefore
preferable to have a clearly defined busbar protection scheme such as unit protection
scheme which will be discussed in the next sub-section.

2.3.2 Unit Protection Scheme

Unit protection scheme is a scheme that operates for a fault within its zone. Here,
zone of protection is decided on the basis of current transformers (CTs), and it includes
each and every fault point inside the CTs where measurement of currents is carried out.
This type of protection scheme is widely used in busbar, generators, transformers, and
large induction motors. Differential and directional comparison protection scheme is the
best example of this type of protection scheme.
2.3.2.1 Directional Comparison
When it is not economical to change out or add new current transformers then bus
fault protection is achieved by directional comparison scheme using the existing line CTs.
The main theme of this scheme is that if the power flow in one or more circuits is away
from the bus, an external fault exists whereas for the power flow in all of the circuits into
the bus, an internal bus fault exists [172]. This scheme is categorized in three parts namely
Series Trip Directional Scheme, Directional Blocking Scheme and Directional
Comparison Scheme.
1) Series Trip Directional Scheme: In this scheme all the directional relay trip contacts
are connected in series. Figure 2.1(a) shows two directional relays R1 and R2 of
line-1 and line-2 for bus protection. Figure 2.1(b) shows its control circuit in which
the contacts of all line relays are connected in series.

BUS + R1 R2 -
BUS Aux
P. T. CB-1 Line-2
AX1 A1
CT-1

R1 CB-2 TC CB-1
AX2 A2

Line-1
CT-2

TC CB-2
R1
(a) (b)
A1 & A2 are mechanical switch, AX1 & AX2 are contacts of Aux relay
TC CB is trip coil of circuit breaker

Figure 2.1 Directional series trip scheme for busbar protection

20
 In case of an internal fault, contacts R1 and R2 closes simultaneously and
energizes the auxiliary relay (Aux) to trip both line circuit breakers. But the several
disadvantages of this scheme are reduction in the reliability of the scheme due to
too many series contacts, complex circuitry and requirement of careful and
periodic review.
2) Directional Blocking Scheme: Figure 2.2 shows the control circuit of the directional
blocking scheme for bus bar protection. The power circuit is same as shown in
Figure 2.1(a). In this scheme, all the tripping contacts (R1 and R2) are connected in
parallel and then to the tripping relay (Aux) whereas all the blocking contacts are
connected in parallel and then to the blocking relay (B). During an external fault,
the blocking relay (B) operates and blocks the operation of bus protection as per
requirement by opening one of its contacts B-1, which is in series with auxiliary
tripping relay. However, the main disadvantage of this scheme is its inapplicability
in large cable network where the capacitance charging current is appreciable in
comparison with minimum ground fault current.

+
R1 -
B
B-1
Aux
R2

A1

AX1 TC CB-1
A2

AX2 TC CB-2

Figure 2.2 Directional blocking schemes for busbar protection

3) Directional Comparison Scheme: The problem of directional blocking scheme can
be solved by this scheme which uses voltage restraint relay. But the main problem
with all types of directional comparison scheme is that the relay settings must be
reviewed and changed whenever system changes are made near the protected bus.
2.3.2.2 Differential Protection Comparison
Protection of busbars is almost universally accomplished by differential protection
scheme. In practice, this type of protection is mainly accomplished by different schemes
such as Circulating Current Differential Protection Scheme, Biased Percentage

21
Differential Protection Scheme, High Impedance Voltage Scheme, Moderately High
Impedance Scheme and Protection using Liner Couplers. Each of this method is discussed
in details below.
1) Circulating Current Differential Protection: Figure 2.3 shows the concept of
circulating current differential protection. This scheme operates on Kirchhoff’s
current law i.e. current entering the substation bus equals to the vector sum of
currents leaving the substation bus. But the main requirement of this scheme is that
the CTs should be of the same ratio or matching CTs are required. However, the
main drawback of this scheme is its maloperation because of production of error
current due to CT saturation and also due to transient DC component.

BUSBAR

CB-1 CB-2 CB-3

Three
CT-1

CT-3
CT-2

R Pole
Relay

Figure 2.3 Circulating current differential protection

2) Biased Percentage Differential Protection: In order to avoid the problem of
maloperaration of relays due to CT saturation and transient DC current, biased
differential protection scheme is used. Figure 2.4 shows connection logic and
tripping characteristic of percentage biased differential protection scheme.

BUS
Operating Current (rms)

CB-1 CB-2
Tripping
CT-1 CT-2 Zone
Blocking
Zone

Restraining
Coil
Operating Coil Restraint Current (rms)

Figure 2.4 Percentage differential protection with percentage-bias characteristic

22
 Application guideline for choosing a secure slope when applying percentage
differential relay for bus protection is given in [132]. Maximum security for external
faults is obtained when all CTs have the same ratio. The main advantages of this
scheme are high tolerance against substantial CT saturation, reduced requirement of
dedicated CTs and its use where comparatively high speed tripping is required. But
the most important limitation of the said scheme is that the relay may maloperate in
case of a close-in external fault due to complete saturation of CT.
3) High Impedance Voltage Scheme: To overcome the problem of spill current due to
CT saturation in case of external fault, the high impedance voltage relaying scheme,
as shown in Figure 2.5, is widely used. The effect of saturation is controlled by
keeping CT secondary & lead resistance low & by adding resistance into relay
circuit. Here, full wave bridge rectifier adds substantial resistance to that leg of
circuit. The series L-C circuit is turned to 50 Hz fundamental frequency in order to
respond only fundamental component of current and make over voltage relay
immune to DC offset and harmonics. A dedicated IEEE standard was assigned as a
guide for protective relay applications to power system buses [81], [43]. This
scheme discriminates between internal and external faults by the relative
magnitudes of the voltage across the differential junction points. The main merits of
the said scheme are stability against transient DC component due to tuned circuit,
improved CT saturation characteristics because of stabilizing resistors and faster
operating time. This scheme, however, is not free from disadvantages. The most
important ones are requirement of dedicated CTs (cost increases), maloperation of
relay in case when the secondary leakage reactance is present and inapplicability of
the scheme to re-configurable busbars.

BUS
CB-1 CB-2 CB-3
CT-1

CT-3

C
CT-2

Variable
L Resistor

O/C
R
Relay

Figure 2.5 High impedance bus differential protection

23
 4) Moderately High Impedance Scheme: This scheme is a combination of high
impedance voltage relay and the percentage differential relay [178]. But the prime
limitation of this scheme is that it requires special type of auxiliary transformer.
5) Protection Using Linear Couplers: Problem of CT saturation in case of iron core CT
is rectified using linear couplers (air core mutual reactors) as they use air core. The
secondary of all linear couplers are connected in series as shown in Figure 2.6. The
output voltage of linear coupler is proportional to the derivative of the input current.
If the voltage sum across relay is zero then input current is equal to output current at
bus. During an internal fault, all line current flows toward bus and thus the induced
voltage appears across the relay. Linear coupler is a special device which requires
low energy relay. This may create problems with change in bus configuration.
However, the need of extra equipments in order to achieve benefits of
microprocessor based relays and high cost are the main drawbacks of this scheme
[148].

BUS

CB-1 CB-2 CB-3

Voltage
LC1

Differential
LC2

LC3

R
Relay

LC – Linear Coupler

Figure 2.6 Bus protection relay with linear coupler

2.3.3 Techniques used in Commercial Relay and their Problems

Digital/Numerical relays provide significant benefits for industrial as well as
commercial systems where an economical but effective busbar protection scheme is
required. These techniques are increasingly being applied to major power system networks
and continuously improving power system operation [156], [129], [120], [37]. Further,
they provide many advantages such as reduction in lifetime management costs,
maintenance cost and also have the ability to extend the protection to cater for primary
system changes with minimal hardware alterations and transmission system changes [95],
[1].
The first numerical busbar protection system was commissioned in 1989. Thereafter,
24
up to the end of 2012, about 500 systems are in operation world-wide with different
busbar configurations [179], [75], [158]. The common practice in busbar protection is to
use numerical busbar protection scheme which provides complete protection for all types
of extra/ultra high voltage bus configurations. The commercially available relays use
innovative techniques such as CT saturation detection in less than 2 ms and dynamic
topology processing algorithms which provides a unique combination of security, speed
(operating time is 15 ms) and sensitivity [150], [36]. At the same time with regard to the
protection of busbars, the reliability and the dependability are the most important
parameters [81], [5]. Moreover, increased in-plant generations heighten the requirement
for fast fault clearing which is required to maintain generator stability [33]. On the other
hand, busbar protection shows a bad track, when comparing the number of times that it has
shown an incorrect tripping (security) in comparison with the number of times that it has
operated correctly (dependability) [6], [93]. Considerations in applying bus protection
schemes to industrial and independent power producer are explained in [157]. Later,
Brewis et al, [98] described busbar protection scheme based on instantaneous elements
contained in numeric overcurrent relays and discussed design issues for busbar protection.
But the failure of instantaneous blocking element in case of external fault is the prime
limitation of the proposed scheme.

2.4 METHODOLOGIES AND COMPARATIVE ANALYSIS

During the last two decade, remarkable success has been achieved in the field of
microprocessor as well as digital/numerical based relaying schemes. Many digital busbar
relaying schemes have been developed by researchers using microprocessors,
microcontrollers, digital signal processors and employing various artificial intelligence
techniques [34], [41]. These are explained in the following sub-sections.

2.4.1 Microprocessor based Busbar Protection Scheme

The block diagram of a typical microprocessor-based relay is shown in Figure 2.7.
The relay has five major components namely analog input system, digital input system,
digital output system, microprocessor unit and storage unit.

25
 Power System Network

CT and PT

Analog Input Digital Output Digital Input

System System System

Microprocessor Based Central Processing Unit

Relay Logic

Storage Power
Unit Supply

Figure 2.7 Block diagram of microprocessor based relay

Analog input system contains signal conditioning devices, analog to digital converter
and filtering circuit. It acquires analog signal (voltages and currents) from the power
system network and convert into digital form. Digital input system acquires data such as
breaker status, isolator status, etc. and provide to the microprocessor unit in digital form.
Digital output system contains digital to analog converter. Storage unit contains Random
Access Memory (RAM), Read Only Memory (ROM) and hard disc storage for permanent
and temporary storage of data. Microprocessor unit samples the data according to the relay
logic and takes decision accordingly.
With the advent of digital technology, researchers and designers have made
remarkable progress in the development of different microprocessor- based algorithms.
Kennedy and his colleague [104] proposed a differential relay for busbar protection based
on harmonic current restraint principle. However, the prime limitation of this scheme is
inhibition of relay operation if the content of harmonics in the differential relay exceeds
the threshold value. Later on Andow and his colleagues [54] presented a low impedance
busbar protection scheme which stabilizes relay performance during CT saturation.
However, the significant disadvantage of this scheme is the inhibition of the relay
operation during CT saturation for a predetermined period. Royle et al. [85] and Wheatley
[93] described a solid-state busbar protection scheme which uses electronic circuits to
identify saturated CTs. But the main drawback of the said schemes is that the differential
relay may not pick-up when electronic circuits bypass the operating coil of the relay for
same portion of the cycle in which the output of a CT is distorted due to saturation. Peck et
al. [49] proposed a new numerical busbar protection scheme using bay units which are
26
located close to the switchgear. The foremost feature of this scheme is that it adjusts
characteristic and restraint factor in order to achieve stability during CT saturation on the
occurrence of heavy through faults. Conversely, delayed operation in case of CT
saturation during an internal fault is the fundamental constraint of the above scheme. Liu
Pei et al. [108] suggested an adaptive busbar protection scheme which incorporates
differential current protection and breaker protection. This scheme accommodates bus
structure change and correctly determines new summations of currents. But when the
currents on the two feeders are same then the separate auxiliary contacts of isolators are
required. Kumar et al. [5] described a differential scheme based on multiprocessing. This
scheme operates with a countermeasure for CT saturation and gives stability against
external fault. But the main drawback of this scheme is that it provides delayed operation
in case of in-zone faults. Furthermore, to achieve fast response in case of internal faults, a
modified scheme has been described which issues a trip signal immediately if the initial
trip decision occurs within a few milliseconds of fault occurrence [205]. However, the
main disadvantage of this scheme is the assumption that for a worst-case internal fault,
none of the CTs will saturate during the first few milliseconds. Thereafter, Kasztenny et
al. [27] proposed a combine busbar protection scheme based on differential and directional
operating principles. This scheme uses duel slop operating characteristic having low and
high slop regions. Usage of differential and directional elements increases security of the
relay. However, the higher operating time of relay during CT saturation on an internal
fault is the prime limitation of this scheme.
2.4.2 Digital Differential Protection Scheme
Digital differential relaying schemes are based on two distinctive architectures:
(i) Distributed busbar protection scheme, as shown in Figure 2.8(a), which uses
Data Acquisition Units (DAUs) installed in each bay to sample and pre-processes
the signals and provides trip rated output contacts. It uses a separate Central Unit
(CU) for gathering and processing all the information and fiber-optic
communications between the CU and DAUs to deliver the data. The main
advantages of this scheme is reduced wiring, however the architecture of this
scheme is less reliable due to complexity in data transfer.
(ii) Centralized busbar protection scheme, as shown in Figure 2.8(b), requires to wire
all the signals to a central location, where a single “relay” performs all the
functions. The wiring cannot be reduced and the calculations cannot be

27
 distributed between a numbers of existing DAUs imposing more computational
demand for the central unit. On the other hand, this architecture is perceived
more reliable and suits better retrofit applications.
BUS
CB CB CB CB CB CB

DAU CT DAU DAU

CT

CT

CT

CT
CT
CU CU
Copper
(a) Fiber (b)
Figure 2.8 Block diagram of digital differential relay
Fast tripping which is an essential requirement in case of busbar faults can be
achieved primarily by low/high impedance differential protection schemes [14], [78]. Till
now, high impedance differential protection scheme is widely used due to its low cost and
simple arrangement. However, the above scheme has several disadvantages such as
requirement of dedicated CTs, high cost due to additional switchgears and unavailability
of numerical protection technology [121]. On the other hand, low impedance differential
protection scheme inherits all the features of digital/numerical protection technology and
can also be applied to different complex busbar arrangements. Kasztenny and his
colleagues [29], [28] proposed low impedance differential protection scheme that
combines percentage differential and current directional protection principles within a
frame of an adaptive algorithm controlled by a dedicated CT saturation detection module.
However, need of large number of inputs and outputs and processing power requirements
are the major shortcomings of the said scheme. These problems can be solved by using
distributed or centralized architectures [66]. But the requirement of large quantities of
specialized hardware is the main weakness of this scheme.
Thereafter, a concept of phase-segregated low impedance digital protection schemes
for medium and large voltage busbars have been proposed in 2001-2002 [26], [76]. This
concept has many benefits such as low cost, initial maturity and simplicity.
Jiali and his colleagues [77] presented a protection scheme based on distributed
principle. This scheme has distinguishing advantages over the traditional centralized
principle such as enhancement in adaptability using the information of the whole power
28
system and reduction in the possibility of a serious blackout of the whole bus due to
maloperation of the busbar protection scheme. However, the main weakness of the above
scheme is the possibility of maloperation in case of severe CT saturation. Wong et al.
[165] described a multi agent system which integrates case based and object oriented
based approaches in the development of a distributed intelligent system. Thereafter, B-Al-
Fakhari and his colleague [19] proposed a differential protection scheme for busbar and
transformer using incremental currents. But the prime limitation of this scheme is that if
CT saturation occurs in the first one or two cycles with an external fault then incremental
currents are not sufficient. Hence, operating quantity exceeds restraining quantity and
relay maloperates.
2.4.3 Performance of Busbar Protection Scheme during CT Saturation and Ratio
Mismatch
The iron core CTs are widely used in the electric power system for the protection of
busbar. These types of CTs are not linear transducers due to the characteristics of the iron
core. Different levels of saturation occur in almost all CTs depending on the magnitude of
the fault current being measured. Figure 2.9 shows a block diagram of CT saturation. The
secondary current I2, transformed from the primary current I1 by a CT, is sampled into a
discrete-time sequence of values, i2(t), by a data-acquisition module. If a sample of i2(t) is
determined within a saturation portion of the fault current waveform by the CT saturation
detection scheme, a compensated sample will be generated by a compensation algorithm.
Finally, the compensated currents are supplied to protective relays. The distorted
secondary current can be compensated based on the detection of the CT saturation period
[71].

Primary
CT
Current
Secondary
Current (I2)
Data Acquisition CT Saturation Normal Protective
(A/D) Detection Relay
i2(t)
Saturated

Current
Compensated

Sample
Current

CT Saturation
Compensation

Figure 2.9 Real-time detection and compensation of CT saturation

29
 The main requirement of busbar protection scheme is to remain stable during CT
saturation [41], [212], [50]. Different researchers have proposed different busbar
protection techniques to improve their performance during CT saturation and different CT
ratio. Sachdev and his associates [127], [72] presented a busbar protection scheme which
estimates the impedances of the positive and the negative sequence circuits in order to
distinguish faults outside the bus protection zone from faults inside the protection zone.
This technique is stable during CT saturation and not affected by CT ratio-mismatch. But
the main disadvantage of this technique is that it requires voltage information which is not
always available in bus protection applications. Thereafter, Eissa [118] proposed phase
comparison based busbar protection scheme. The above two schemes are based on phase
angle comparison and provide improved results in case of CT saturation and CT ratio
mismatch than the algorithms based on magnitude comparison. However, the prime
limitation of the above two techniques is the delayed operation in case of severe CT
saturation. Kang and his colleagues [203], [202] presented a CT saturation detection
algorithm based on the detection of start and end of each saturation period using a third
difference function applied to the secondary current signal. The above two schemes ensure
stability in case of external fault with CT saturation. However, the main drawback of the
said two schemes is the difficulty in case of a progressive fault as the second fault may be
detected as the saturation start or end. Guzman and his co-workers [16] proposed a reliable
busbar protection scheme with advanced zone selection that is secure for external faults
with heavy CT saturation. But the fundamental demerit of the above scheme is that the
relay requires primary CTs that reproduce the primary current without saturation for at
least 2 ms after the inception of external fault. In order to further improve the performance
of the busbar protection scheme, Kang and his colleagues [204], [205], [201] suggested a
compensating algorithm for the distorted current signal of a saturated CT. This scheme is
unaffected by CT saturation and successfully operates even for a progressive fault from a
feeder fault to a busbar fault. However, the stability of the proposed scheme cannot be
guaranteed during severe CT saturation and it is also affected by CT error and CT ratio
mismatch.
2.4.4 Neural Network Based Busbar Protection Scheme
Artificial Neural Networks (ANN) are biologically inspired and consist of elements
that are organized in way that may be related to the anatomy of the human brain. These

30
elements called neurons are capable of realizing basic logic functions. The unique feature
of ANN structure can be used to estimate any continuous function.
Figure 2.10 shows the block diagram of Artificial Neural Network (ANN) based
differential relaying scheme for busbar protection. Current samples in secondary form,
acquired from the power system network are given as an input to Analog to Digital
Converter (ADC) through Signal Conditioning (SC) and Anti Aliasing Filter (AAF) block.
AAF is a low pass filter and used to remove high frequency signals. The differential
current signals are given as an input to ANN block. Depending upon nonlinearity in the
input signal, the structure of ANN can be decided. Generally, the structure of ANN
contains input layer, output layer and two or three hidden layers. The output of ANN is
given to decision logic block, which gives tripping or blocking signal.

AAF
CT Decision
Power System SC Trip
Units ANN Logic
Network Id ADC

Figure 2.10 Block diagram of ANN based busbar protection scheme

In the last two decades, the use of ANN for solving problems in power system is
rapidly increasing [189]. Feser and his colleague described [100] ANN based busbar
protection scheme in which a neural network is used for preprocessing the data and
restoring the distorted signals. This work is further extended on a larger data base with an
increased network size [187]. But the CT error and CT ratio-mismatch can still cause the
maloperation of the proposed scheme. Thereafter, Zadeh and his co-worker [122], [69]
proposed Fuzzy Neuro based pattern classifier scheme which distinguishes faults in a
busbar protection zone from those outside the zone. The main advantage of the above
scheme is the stability during CT saturation.
But, large training sets, tedious training process, and a large number of neurons are
the several disadvantages of the neural network based schemes. Moreover, another
significant weakness of neural-network-based algorithms is that the training may not
converge in some cases, as the starting point is chosen randomly and can end up in a local
minimum. This is quite common, particularly in the case of retraining the neural network
with new data associated with, for example, contingencies that may not converge to the
desired value. When the learning gets stuck on local minima, the requisite performance
can never be reached. Thus, the neural networks still remain black boxes and lack
transparency [181].
31
2.4.5 Travelling-Wave Based Busbar Protection
The travelling wave based protection scheme measures the relative time of arrival of
the travelling wavefront produced by the fault. The traveling wave detection system
consists of Capacitive Coupling Voltage Transformer (CCVT), Fault Transient Interface
Unit (FTIU), Global Positioning System (GPS) receiver with event storage and master
station with software. Based on this logic, researchers have given different schemes which
are summarized below.
Wang and his co-workers [61] presented a busbar protection scheme based on
polarities of transient current waves for identification of the in-zone or out of zone faults.
Jiang and his colleagues [56] proposed Transient Based Protection (TBP) technique which
depends on energy spectrum of fault transient. Subsequently, Chen and his co-workers
[210] described busbar protection scheme based on wavelet analysis which works on fault
generated transient current signals. Afterwards, Eissa [117] proposed a directional
technique based on a combined signal derived from pre-fault voltage and post fault current
signals. Subsequently, Xiang-ning et al. [199] presented directional relaying scheme for
the protection of transmission line and busbar based on the polarity comparison of the
current traveling waves. However, when a fault occurs at a voltage inception angle close
to zero degrees, it does not generate significant traveling wave components. In addition,
the bandwidth limitation of the transducers, particularly the CVT, limits the applicability
of such techniques, particularly for close-up faults, which generates very-high-frequency
traveling waves that are outside the bandwidth of receptibility of conventional CVTs [21].
2.4.6 Wavelet Transform Based Busbar Protection Scheme
Figure 2.11 shows the block diagram of Wavelet Packet Transform (WPT) based
differential relaying scheme for busbar protection. Current samples, in secondary form, are
received from the power system network. Thereafter, they are given as an input to Analog
to Digital Converter (ADC) through Anti Aliasing Filter (AAF) and Signal Conditioning
(SC) device. Afterwards wavelet packet transform block decomposes the differential
current signal of the respective phase into detail and approximation coefficients. These
coefficients are used in relay logic block for final decision about internal and external
fault.

32
 CT

Id

AAF Decomposition of
Relay Fault
SC different current Trip
Logic Classification
ADC Using WPT

Figure 2.11 Block diagram of wavelet based busbar protection scheme

Digital protective relaying of transmission lines has been greatly benefited from the
development of artificial intelligence and signal processing techniques [154]. Fernandez et
al. [160] presented an overview of the wavelet transform applications in a power system,
which indicates that the use of wavelet transform in power system protection is increasing
day-by-day. Eissa [118] presented busbar protection scheme which uses Continuous
Wavelet Transform (CWT) in order to derive operating and restraining quantities. Soon
after, Mohammed [112] described differential busbar protection technique using Wavelet
Packet Transform. However, the above two schemes may maloperate in case of varying
load conditions. Thereafter, Valsan et al. [169] presented a new busbar protection
technique based on directional signal which is derived as the product of high frequency
details of branch currents and bus voltage using a Wavelet transform approach. But the
main disadvantage of this technique is that it requires voltage information which is not
always available in bus protection applications.
2.5 CONCLUSION
In this chapter an attempt is made to review most of the existing schemes of busbar
in power system networks. The chapter also covers the concepts of the different protection
schemes of busbars, with a bibliographical survey of relevant background, practical
requirements, the historical events, the present state and techniques. Comparative
evaluation of different techniques is also undertaken for identification of optimum
technique as required by specific application. The citations listed in this bibliography
provide a representative sample of current engineering thinking pertaining to next
generation busbar protection problem. This chapter is based on many research articles
published in the last 50 years and periodic bibliographic updates on this topic will be
useful as the restructuring of power system network continues to evolve. Although,
33
numbers of busbar protection schemes have been proposed so far there exists a lot of
scope for further improvement especially on the effective discrimination between in-zone
and out of zone fault. A clear consensus is presently heading towards SVM and other AI
based digital busbar protection scheme and can handle the present day complex busbar
protection problem.

34