Using advanced distribution
protection functions can
significantly reduce faultclearing times.
The electric power industry is going through
significant changes due to the increased requirements for
improved quality of power supplied by the utility in order to
avoid costly interruptions of manufacturing or other processes
caused by voltage sags, swells or unbalanced conditions when
a short circuit fault occurs in the distribution system.
The improvement of power quality during short circuit faults
can be achieved in several different ways. Like any other
problem that has to be solved, we need first to understand
the nature of the problem and its effect on sensitive users.
The most common short circuit faults in the system –
single-phase to ground faults – are characterized by the fact
that they introduce a voltage sag in the faulted phase, and at
the same time they result in a voltage swell in the two healthy
phases. This is clearly seen in Figure 1 that shows the recorded
waveform for a single-phase to ground fault.
The effects of voltage sags and swells on sensitive
equipment have been studied for many years by industry
organizations such as Computer and Business Equipment
Manufacturers' Association (CBEMA). They record both
characteristics of this power quality event – the depth of
the sag and its duration. Figure 2 shows a plot of depth
vs. duration known as the ITIC (Information Technology
Industry Council) or CBEMA graph.
The first characteristic of voltage sags – the depth – is a
function of the type of fault, fault location and the system
configuration. This is something that we can not control.
The second characteristic of voltage sags – the duration – is
the parameter that we can control by applying the advanced
features of multifunctional protection relays. Monitoring the
changes of the power system configuration and adapting the
relay to these changing conditions can further improve the
performance of the relays and reduce the effect of short circuit
faults on sensitive equipment or processes.
The effects of short circuit faults on the voltage profile
across the distribution system may result in the shutting
down of sensitive industrial processes. The performance of
typical distribution feeder or substation protection systems
historically has not been considered as critical as the behavior
of transmission line protection devices. This is changing now,
resulting in a new look at the requirements and functionality
of distribution protection relays.
The effect of changes of the system configuration on the
performance of the protective relays is another factor that
needs to be considered. Using all available protection and
programmable logic functions in multifunctional protective
IEDs can help us significantly reduce the effect of short circuit
faults on sensitive loads supplied from the distribution
substation. Adaptive protection based on detected changes
in the system configuration, combination of instantaneous,
definite time and inverse time-delayed phase, ground and
negative sequence elements, will shorten the fault duration,
resulting in changes in the voltage level/time characteristics
of the fault condition and reduced probability for the costly
interruption of voltage sensitive processes.
Optimizing distribution protection
Typical distribution feeder protection is based on phase
and ground overcurrent relays set to protect the line for
three-phase, phase-to-phase or phase-to-ground faults.
An instantaneous relay is used to operate for close-in faults
and a time overcurrent relay with inverse characteristic
provides protection for most faults on the line. The time
overcurrent relay has to coordinate with any fuses used to
protect distribution transformers connected to the feeder.
The coordination requirements for high current faults result
1 Single phase-to-ground fault
2 Inverse-time characteristics
U/V
1000
50
100
-
0
I
I
I
I
I
I
I
I
-100 -
100
-50
10
VA-L2
VA-L3
10
2-
4
0
-2
I
I
I
I
I
I
I
-
IEC SI
1
RECT
IEC VI
1
IEEE MI
IEEE VI
IEC EI
IA-L1
PAC.SPRING.2009
I
UK LTI
Operating Time (s)
VA-L1
I/V
Operating Time (s)
Distribution Protection
cover story
20
IA-L2
IA-L3
0.1
1
10
100
0.1
1
10
US CO2
US CO8
IEEE EI
100
by Damien Tholomier, AREVA T&D Automation, Canada
21
3 ITIC (CBEMA) curve - Revised 2000
Percent of Nominal Voltage (RMS or Peak Equivalent)
500
400
300
Prohibited Region
Voltage Tolerance Envelope
Application to Single-Phase
120-Volt Equipment
200
140
120
100
No Interruption Function Region
80
70
110
90
No Damage Region
40
0
1 us
0.001 c
0.01 c
1c
1c
1 ms 3 ms 20 ms
100 c
0.5 s
10 s
Steady
State
Duration in Cycles (c) and Seconds (s)
4 Inverse versus definite time overcurrent
10
Seconds
in increase in the operating times for faults further down the
line, with the longest times for faults at the remote end.
In order to reduce the number of electromechanical or solid
state relays, backup protection for bus faults or breaker failure
has been traditionally provided by the transformer protection
relays. Considering the fact that they also have to coordinate
with the feeder relays, it is obvious that the operating times
for bus faults or feeder faults with breaker failure will not
meet the requirements of sensitive customers.
Modern multifunctional protection relays have many
features that allow significant improvements in the
performance of the relays under different fault conditions.
Definite Time versus Inverse Time Overcurrent
Modern distribution feeder protection relays have multiple
phase and ground overcurrent elements used to reduce the
operating time of the relay for different fault conditions.
Multiple independent stages are available for each phase
overcurrent element and the two ground fault elements - one
based on measured quantities and the second - on derived
quantities calculated from the three phase currents.
Each stage may be selected as non-directional or
directional (forward/reverse). All stages usually have definite
time delayed characteristics and some of the stages may also
be independently set to inverse-time IEC or IEEE tripping
characteristic (Figure 2). These stages have a programmable
reset timer for coordination with other devices.
A voltage controlled overcurrent function can be enabled
on phase overcurrent elements. It provides back-up protection
for remote phase faults whilst remaining insensitive to load.
The directionality of the ground fault elements is provided
by residual voltage or negative sequence voltage polarizing.
A separate multi-stage sensitive ground fault element is
provided and is selected as non-directional or directional.
Figure 4 shows the inverse time characteristic of a phase
overcurrent relay with the operating points for different
fault locations on the protected feeder. The characteristic is
coordinated with a downstream fuse.
Table 1 shows a comparison of the operating times of the
traditional inverse time characteristic, compared with the
use of the two additional definite time delayed overcurrent
elements with settings of 0.2 s and 0.35 s.
The last column in the table gives the difference in
the operating times. The advantages are quite significant,
especially for the faults closer to the end of the line.
Negative Sequence Overcurrent Protection
Overcurrent elements in traditional phase overcurrent
protection must be set higher than maximum load current.
This limits the sensitivity of the phase relays and at the same
time results in increased operating times for line end faults.
Since the levels of the phase fault currents for phase-to-phase
faults are lower than the levels for three-phase faults, this will
result in further increase of the fault-clearing time.
If we consider an example for a phase-to-phase fault at 75
% of the protected feeder, the voltage at the relay location is
0.79 Unom and the phase overcurrent relay with inverse
characteristic used for Table 1 and Figure 4 will operate in
07
2. CIRED PTOCph MCG G-EI TD=0.500
CTR=100 Tap=0.25A(Dir) Inst=1000A
TP=1.6667s
05
04
03
02
1
0.7
0.5
0.4
1. CIRED Fuse
153-151-050
Idiv=0.3 Minimum melt.
0.3
0.2
0.1
0.07
0.05
0.04
0.03
0.02
0.01
10
2
3
4 5
7
100
2
3
4 5 7
1000
Current (A)
PAC.SPRING.2009
Damien Tholomier received a
BEng in Electronic,
Electrotechnique
and Automation in
1990, and a Master
in Electrical and Automation Engineering in 1992 from the
University of
Marseilles,
France (Ecole
Polytechnique).
He joined ALSTOM
T&D GmbH in
Stuttgart, Germany
where he worked
for 5 years in
the Protection &
Control department.
In 1997 Damien
moved as Marketing Manager High
Voltage Protection
Business Unit with
Alstom T&D Protection & Control in
Lattes, France.
From 2002-2006
he was Marketing Protection
Relays Director for
ALSTOM later AREVA
T&D Automation.
Damien is currently
Managing Director
in Canada, and he is
also responsible for
Protection and
Automation in
North America.
Damien is CIGRE,
IEEE, IEC TC95 and
GIMELEC member.
22
Distribution Protection
cover story
0.71 s. For the same three-phase fault the voltage drop seen
by the relay is 0.71 Unom with operating time of 0.53 s.
Any unbalanced fault condition will produce negative
sequence current of some magnitude. Thus, a negative
sequence overcurrent element can operate for both
phase-to-phase and phase-to-ground faults. If a definite time
negative sequence overcurrent element is used, it may be
set to clear the same phase-to-phase fault in less than 0.3 s.
This will bring the duration of the fault within the acceptable
region of No Interruption in Function. Negative sequence
table 1 Operating Time Comparison
Fault Location
TOC Time
DTOC Time
Time
40 %
50 %
60 %
70 %
80 %
90 %
100 %
0.24 s
0.31 2
0.39 s
0.48 s
0.59 s
0.70 s
0.82 s
0.20 s
0.20 s
0.35 s
0.35 s
0.35 s
0.35 s
0.35 s
0.04 s
0.11 s
0.04 s
0.13 s
0.24 s
0.35 s
0.47 s
Distribution systems present different challenges to the prote
overcurrent elements also give greater sensitivity to resistive
phase-to-phase faults or high resistance phase-to-phase-toground faults, where phase overcurrent elements may not
operate at all.
RMS Thermal Overload
Thermal overload protection based on a thermal replica
using RMS load current to model heating and cooling of the
protected equipment can be used with both alarm and trip
stages. The heat generated within a cable or a transformer
is directly proportional to current squared. The thermal
time characteristic used in the relay is therefore based on
current squared, integrated over time. The largest phase
current should be used for input to the thermal model. The
thermal element may be set with either a single time constant
characteristic for the protection of cables or dry transformers,
or a dual time constant characteristic to protect oil filled
transformers. See the characteristic in Figure 5 bellow.
Modern
distribution
protection
relays can be
used not only
5 Dual time constant thermal overload
to clear short
circuit faults,
characteristic
10000
0
protect the
equipment
under
emergency load
conditions.
Operating Time (seconds)
but also to
1000
0
Time constant 1 = 5 mins
Time constant 2 = 120 mins
Pre-overload current = 0.9 pu
Thermal setting = 1 Amp
100
0
10
0
1
0
1
1
PAC.SPRING.2009
Current as a Multiple of Thermal Setting
1
0
Distribution Bus Protection
The selection of protection equipment for bus faults until
recently was based on the requirements for stability of the
power system. Because of that the protection of buses in the
case of short circuit faults at the transmission level is usually
provided by high or low impedance bus differential relays.
The distribution bus protection has been done by the backup
time delayed overcurrent protection of the transformers.
The understanding of the effects of longer fault clearing
times on sensitive industrial equipment results in a change
in the philosophy on distribution bus protection. It is now
based on exchange of signals between the feeder relays and
the transformer protection relays.
All overcurrent starting signals from the multiple feeder
relays are paralleled and used to energize an opto-input of the
transformer overcurrent protection relay. For a fault on any of
the distribution feeders, the relay protecting the faulted feeder
will start and with or without time delay (depending on the
fault location) will issue a Trip signal to clear the fault.
Figure 6 shows a simplified block diagram of a distribution
bus protection. If the fault is on one of the feeders, the
protective relay of that feeder will immediately operate an
output that is wired to an input of the transformer relay. This
signal will indicate a feeder fault and will block the operation
of the bus protection function.
If the fault is on the bus, no feeder relay will operate, thus
indicating to the transformer protection relay that it is a bus
fault. The overcurrent elements that are used to implement
a distribution bus protection scheme have to be set with a
certain time delay that allows the receiving of a signal from
any of the feeder relays. At the same time each feeder relay
should be able to communicate the starting of an overcurrent
element that is used to block the bus protection element.
The advantage of this type of scheme therefore is that it
allows fast fault clearing of distribution bus faults without the
23
Multifunctional distribution
protection devices can adapt to
changing system conditions.
Unbalanced
loads and
line-end
phase-tophase faults
require
ection system.
the use of
negative
sequence
overcurrent
need for installation of a distribution bus
differential protection.
protection.
Breaker Failure Protection
The requirements for improvements in the quality of
power supplied to electric utility customers result in changes
of the way distribution feeder protection is designed and
applied. Many protection schemes that in the past have only
been used at the transmission level today are common at the
distribution level. One of the reasons is that they are available
as some of the numerous functions in a multifunctional
distribution feeder relay. One of these schemes is the Breaker
Failure Protection. Breaker failure results in prolonged
exposure of industrial customers to low voltages and of
electrical equipment to large short circuit currents, and may
lead to damage of equipment and complete shut down of the
manufacturing process. This is the reason that Breaker Failure
Protection has gained popularity at the distribution level of
the system.The most common Breaker Failure Protection is
based on monitoring of the current in the protected circuit.
After a fault is detected and the relay issues a trip signal, it
will also initiate the timer of the Breaker Failure Protection
function. If the breaker trips as expected, the current in all
three phases will go to zero, which will reset the undercurrent
element used to detect the correct breaker operation. Since at
the distribution level the feeder is protected by a single relay,
the Breaker Failure Protection function is usually started by a
built-in protection function in the protection relay.
Fuse Saving Scheme
The problem with fuse protection of distribution
transformers is that it does not allow automatic restoration of
the power supply and requires a crew to be sent to the location
to replace the fuse, leading to long supply interruption. Since
most short circuit faults have a temporary nature, attempting
to clear the fault before the fuse burns has become a standard
practice in many utilities using a Fuse Saving Scheme. It uses a
low set instantaneous overcurrent element to trip the breaker
in the substation immediately after the fault occurs. There is
no coordination of the instantaneous overcurrent element
with the downstream fuses. The breaker is tripped before the
fuse protecting the faulted element will start to melt. After the
reclosing, the low set instantaneous element is disabled and
overcurrent elements that coordinate with the downstream
protective devices are used.
The advantage is that in case of a temporary fault the fuse
is not going to melt, i.e. it will not require a replacement and
will result in a short interruption of the load during the dead
interval of the reclosing sequence. This can be very important,
especially in cases where the fuse is at a remote location.
The disadvantage is that all the customers supplied from
the feeder will be affected by the interruption during the
reclosing cycle. That is why the decision to apply the Fuse
Saving Scheme should be made based on the sensitivity of the
loads to voltage sags or interruptions.
Selective Backup Tripping
Protection of distribution feeders today is commonly
provided by a single multifunctional relay. If the relay fails
and at the same time there is a fault, the protection is typically
provided by time-delayed overcurrent elements of the
transformer protection that trips the transformer breaker.
This has the negative result of first, delayed operation and
second, the tripping of the source breaker leading to a voltage
interruption of all feeders connected to the distribution bus.
A significant improvement can be achieved by the selective
backup tripping from the transformer relay. If it receives a
signal that the feeder relay has failed, when a fault is detected
and there is no blocking signal from any of the healthy feeder
relays, the transformer relay will first send a trip signal to the
breaker of the feeder with the failed relay. If the fault is on that
feeder, it will be cleared, thus eliminating the need for tripping
the transformer breaker and causing the voltage interruption
for all feeders.
Load shedding in substations
Load shedding in substations can be executed based
on several different principles. It also can be triggered by
different system events and can serve different purposes. Load
shedding is executed locally in the distribution substations
and triggered by the drop of frequency below the pre-defined
thresholds in accordance with the defensive plans. Load also
can be shed in order to prevent the separation of the system.
It is triggered by criteria implemented in a system integrity
PAC.SPRING.2009
Distribution Protection
cover story
24
Advanced
Advanced load-shedding
functions are available
in modern distribution
protection devices.
protection scheme (SIPS) and executed in substations when
they receive commands for load shedding from the SIPS.
Distributed load shedding is a relatively new concept since
it requires each individual distribution feeder to be equipped
with an IED that measures the frequency and can perform
the load shedding function. This means also that each of the
feeder relays need to have voltage inputs. Such a device also
provides measurements, recording and other functions.
In case of distributed load shedding each individual
relay belongs to a specific step of the load shedding system
and usually has a more limited functionality compared to
specialized IEDs used in centralized systems.
The fact that each feeder can be controlled by a separate
step in the load shedding system with a different setting,
allows the implementation of a more flexible system that will
shed load closer to the requirement for balancing load and
generation in the area separated after a disturbance.
Load shedding functions in multifunctional protection
IEDs can be achieved using different methods and their
combination in complex schemes using the programmable
scheme logic that such devices typically have.
Underfrequency The basic underfrequency protection is the most
commonly used and is available with multiple independent
definite time delayed stages – up to six in some relays. The
elements are usually definite time delayed, but can also be
used as instantaneous.
In order to minimize the effects of underfrequency on
a system, a multi stage load shedding scheme may be used
with the substation loads prioritized and grouped. During an
underfrequency condition, the load groups are disconnected
sequentially depending on the level of underfrequency, with
the highest priority group being the last one.
Frequency supervised rate of change of
frequency
Considerable load to generation imbalance may result in
relatively rapid changes of the system frequency. In such a
case maintaining the system stability is an important task, and
requires quick corrective action. High speed load shedding
cannot be achieved by monitoring the system frequency
alone and the rate of change of system frequency becomes an
equally critical parameter to use.
Independent rate of change of frequency
This element is a plain rate of change of frequency
monitoring element, and is not supervised by a frequency
setting. It can be used to provide extra flexibility to a load
shedding scheme in cases of severe load to generation
imbalances. Since the rate of change monitoring is
independent of frequency, the element can identify frequency
variations occurring close to nominal frequency and thus
provide early warning to the operator on a developing
frequency problem. Additionally, the element could also
be used as an alarm to warn operators of unusually high
frequency variations.
Average rate of change of frequency
Variations in frequency during times of generation – load
imbalance do not follow any regular patterns and are highly
non-linear. Oscillations will occur as the system seeks to
address the imbalance, resulting in frequency oscillations
typically in the order of 0.1Hz to 1Hz, in addition to the basic
change in frequency.
The rate of change of frequency elements use an
“instantaneous” measurement of “df/dt”. Due to the
oscillatory nature of frequency excursions, this instantaneous
value can sometimes be misleading, either causing unexpected
operation or excessive stability. For this reason, the relays
provide an element for monitoring longer term frequency
trend, thus reducing the effects of non-linearity in the system
and providing increased security. It is recommended that
the average rate of change of frequency element is used in
conjunction with the frequency element.
6 Diagram of distribution bus protection
7 Distributed load shedding system
Substation
I Fdr1
protection
I x-er
schemes allow
I Fdr2
I Fdr3
Trip
significant
reduction in the
TOC
Start
fault clearing
Timer
AND
Blocking
Logic
time.
HV
Load
IOC
Start
PIED 1
P Fdr1
PAC.SPRING.2009
PIED 2
P Fdr2
PIED i
P Fdri
25
Programmable scheme logic
Modern state-of-the-art numerical relays incorporate
many protection and control elements. These elements
can be used as building blocks to configure the device to a
customer’s specific application.
The Programmable Scheme Logic is the means by which
the internal elements, in reality numerical algorithms, are
interlinked in accordance with conventional gate logic rules.
The PSL also assigns how opto inputs assert control over the
enabled elements, and which outputs should change state,
should a particular combination of elements operate. The
PSL is used to supplement built-in protection and control
schemes, such as breaker-failure protection, which are
already implemented within the standard building blocks of
elements. Some of the benefits offered by multifunctional
devices configured in PSL include:
Easy configuration of customer-specific schemes
Protection reconfiguration, when required
Removing the need for a programmable logic controller
and circuit breaker bay controller.
The user can design a logic scheme based upon three main
types of logic gates: logical AND, OR, and MAJORITY gates.
Any gate input or output can be inverted, allowing NOT,
NAND, and NOR configurations. Auxiliary timers may
be assigned at any stage of the logic for sequence of events
coordination, also within feedback loops if required.
Event driven PSL only writes to output contacts or
front plate LEDs once all PSL re-evaluation is complete.
The principle of event driven PSL is to reduce the quiescent
overhead on the processor to a very low level, and to keep
it low even during fault and autoreclose conditions on the
power system. This ensures that fast and repeatable trip times
are obtained from relays, and that non-real time tasks such as
fault location and disturbance record management take place
immediately afterwards.
Monitoring
Predictive maintenance is becoming extremely important
in the efforts of utilities to deal with reduced personnel and
increasing customer requirements for improved power quality
and reliable supply of electric power. Modern distribution
protection IEDs measure and calculate numerous analog
parameters and provide additional monitoring functions
that allow the transition from scheduled to event driven
maintenance, without the need for specialized equipment.
A variety of tools, available in microprocessor based
relays can help the user to determine the need for primary or
secondary substation equipment maintenance based on user
defined alarm signals from the protective device.
The data is divided in several categories:
Breaker maintenance related data, including breaker
interrupted current, breaker operation counters, fault counters
and breaker opening and close time monitors
Breaker auxiliary contacts and trip coil monitoring
Voltage transformer supervision schemes
Current transformer supervision logic
Broken conductor detection
All the above information is available from the relays based
on built in analysis tools that process the fault data and convert
it to information, thus reducing the need for protection,
control and maintenance personnel in the decision process.
Recording
The relays need to record and time tag different events
and store them in non-volatile memory. This enables the
system operator to establish the sequence of events that
occurred within the relay following a particular power
system condition, switching sequence etc. The real time clock
provides the time tag to each event, with a resolution of 1ms.
The event records consist of fault flags, pre-fault and fault
measurements etc. Fault location is part of the data included
in the fault record.
Internal failures detected by the self-monitoring circuitry,
such as watchdog failure, field voltage failure etc. are logged
into a maintenance report.
Recording of the waveforms during a distribution system
event can help with the analysis of the operation of the
protection relays. The number of records that may be stored by
the relay is dependent on the selected recording duration. The
duration of the recording may need to be extended to several
seconds in order to capture the operation of time-overcurrent
protection elements.
8 Average rate of change of frequency
9 Programmable scheme logic
f
Supervising frequency
True slope for the time
t
F
f
t
t
PAC.SPRING.2009
Programmable
Scheme Logic
developed using
user-friendly
engineering
tools allows the
implementation
of application
specific
protection
schemes .