Siemens - Applications For SIPROTEC Protection Relays

2005
Applications for
SIPROTEC Protection Relays
Contents
Line Protection in Distributions Systems Page
Protection of Combined Cable and Overhead Lines 3
Redundant Supply with Bus Coupler 7
Coordination of Inverse-Time Overcurrent Relays
with Fuses 11
Medium-Voltage Protection with Auto-Reclosure
and Control 21
Differential Protection of Cables up to 12 km via
Pilot Wires (Relay Type: 7SD600) 31
Applications for Differential Protection of Cables via Fiber Optics
(Relay Type: 7SD610) 37
SIPROTEC Protection Relays Thermal Overload Protection of Cables
Disconnecting Facility with Flexible Protection
47
Function 53
2005 Earth-Fault Protection in Systems with Isolated
Star Point 59
Earth-Fault Protection in a Resonant-Earthed System 63
Earth-Fault Protection in a Low-Resistance-
Earthed System 67
Coping with Single-Phase Load Diversity Using
Adaptive Relay Settings 69
Line Protection in Transmission Systems

400 kV Overhead Transmission Line Protection 73
Directional Phase Overcurrent Protection ANSI 67
with 7SA522 and 7SA6 117
Distance Protection with Parallel Compensation 121
Protection of Long Lines with SIPROTEC 7SD5 135
Transformer Protection
Protection of a Three-Winding Transformer 143
Protection of a Transformer with Tap Changer 153
Protection of an Autotransformer 159
Motor Protection
Protection of a Motor up to 200 kW 169
Generator Protection
Protection of a Medium-Sized Generator up to 5 MW 179
System Solutions for Protecting Medium and Large
Power Station Units 187
Protection of Medium-Sized and Large Generators
with SIPROTEC 7UM6 193
Unit Protection System for Pumped-Storage Power
Stations 207
Busbar Protection
Application of Low-Impedance 7SS601 Busbar
Differential Protection 213
Basic Busbar Protection by Reverse Interlocking 223
© Siemens AG 2005
Line Protection in Distribution Systems
Protection of Combined
Cable and Overhead Lines
n 1. Distance protection with auto-reclosure on

mixed lines
On mixed lines with cables and overhead lines the
distance zone signals can be used with a distance
protection relay 7SA6 to distinguish to a certain
extent between cable and overhead line faults.
Mixed lines mean that part of the protected line is
designed – within the same grading zone – as a
cable section and the other part of that line as an
overhead line section. The auto-reclosure function
LSP2318.eps
is only useful on the overhead section of the line.
The section of the line to be protected must be
selected accordingly in the grading. The auto-rec-
losure function can be blocked (in the event of a
fault in the cable section) by interconnection by
means of the user-programmable logic functions
(CFC) in the DIGSI parameterization and config- Fig. 1 SIPROTEC 7SA6 distance protection
uration tool.
n 2. System configuration
According to the system configuration in the dis-
tance zones Z1, Z2, Z3 and Z5 with their line im-
pedances (impedances of the line as R and X va-
lues, resistance values and reactance values) the
line sections are graded as usual in the distance
protection relay SIPROTEC 7SA6. Zone Z1B
serves above all for the automatic reclosing func-
tion and for switching functions (e.g. “manual
close”). Zone Z4 is used to measure and select the
Fig. 2 Protection of combined cable and overhead lines
cable or overhead line part of the line to be pro-
tected.
The application is described on the basis of a line
Zone Z1B can also be used for fast disconnection
<A-B> with two SIPROTEC 7SA6 distance pro-
of the line to be protected when closing onto a
tection relays.
fault, in addition to application in conjunction
with the auto-reclosure function. The protection A solution for the protection relay at location “A”
must trip fast if, when closing onto the line to be is described in the following. Here the line sec-
protected, the feeder at the remote end is e.g. still tions of the mixed line are selected with the dis-
earthed. The 7SA6 also provides the “high-current tance zones Z1B and Z4.
– instantaneous tripping” function for this protec -
The protection relay at location “B” can be set for
tion as an alternative. Both applications are de-
auto-reclosure on mixed lines either
scribed here.
n by the distance zone Z1B and the high-current
instantaneous tripping or
n alternatively with grading of zones Z1B and Z4.
Siemens PTD EA · Applications for SIPROTEC Protection Relays · 2005 3

n 3. Settings in the configuration with DIGSI n 4. Creating the logic flowcharts

First the following settings must be made in the All that need now be done is to create, link and
configuration matrix in the parameter set for the translate the appropriate logic diagrams in the
7SA6, for configuration in DIGSI. CFC in DIGSI. The “fast” PLC task (PLC0) is used
as a run level in the CFC.
Configuration matrix (group: “Auto-reclosure” or
“General distance protection” The individual logic functions and the effect on
the protected zone are described below.
a) FNo. 2703 – “>Block auto-reclose function”
configured to “Source CFC” Appropriate allocations must be performed (with
the execution of an auto-reclosure function) for
b) FNo. 3747 – “Distance pickup Z1B, loop L1E”
the described line <A – B> in both distance pro-
configured to “Target CFC”
tection relays, to detect the zone of the overhead
c) FNo. 3748 – “Distance pickup Z1B, loop L2E” line.
4.1 Control of auto-reclosure in the 7SA6 for
d) FNo. 3749 – “Distance pickup Z1B, loop L3E”
protection relay A
e) FNo. 3750 – “Distance pickup Z1B, loop L12”
f) FNo. 3751 – “Distance pickup Z1B, loop L23”
g) FNo. 3752 – “Distance pickup Z1B, loop L31”
h) FNo. 3759 – “Distance pickup Z4”
configured to “Target CFC” Fig. 3 Control of the auto-reclosure for protection relay A
Parameterization: (parameter group A, distance

protection – polygon, zone 4) 7SA6 – protection relay A:
Parameter 1335 “T4 DELAY”
The setting values of zone Z4 correspond to the
The tripping time for zone 4 (parameter 1335 = grading with the R and X values of the cable sec-
T4) must be set to infinite (T4 = ∞) because this tion. Zone Z1B is designed as usual for about
zone is only used for selecting the cable or over- 120 % of the line length. Since no auto-reclosure
head line part of the line. Zone 4 should only sig- is to be performed in the cable section, the over-
nal a pickup in this application. Tripping in this head line section is selected in zone Z1B by a CFC
zone is irrelevant. This function is particularly im- plan. With the result of the CFC plan (FNo. 2703:
portant for the single-pole auto-reclosure func- “>AR block.”) auto-reclosure can be blocked in
tion because the tripping then takes place the event of a fault in the cable section (zone Z4)
exclusively via zone Z1B. (see Fig. 4).
>AR block
(FNo. 2703)
LSP2689en.tif
Fig. 4 CFC plan for protection relay A
4 Siemens PTD EA · Applications for SIPROTEC Protection Relays · 2005

4.2 Control of auto-reclosure in the 7SA6 for At least 2.5 times the rated current of the line is
protection relay B recommended as a pickup value for the instanta-
neous high-current tripping.
4.2.1 Version 1 4.2.2 Version 2
Fig. 5 Distance protection with zone Z1B and the instan- Fig. 6 Distance protection with grading of the zones Z1B
taneous high-current tripping and Z4
7SA6 – protection relay B: 7SA6 – protection relay B:
The setting values of zone Z1B correspond to the The setting values of zone Z4 correspond to the
grading with the R and X values of the overhead grading with the R and X values of the overhead
line on which the auto-reclosure function is to be line. Zone Z1B is designed as usual for about 120 %
performed. The instantaneous “high-current trip- of the line length. Since no auto-reclosure is to be
ping” function is used in 7SA6 for instantaneous performed in the cable section, the overhead line
tripping when closing onto a fault, to completely section is selected in zone Z1B with a CFC plan.
protect the <A – B> line. With the result of the CFC plan (FNo. 2703: “>AR
block.”) automatic reclosing is blocked in the
The task of the instantaneous high-current ele- event of a fault in the cable section. This means
ment (instantaneous high-current switch-onto- that an auto-reclosure is only performed in the
fault) is to perform tripping immediately and case of pickup of the protection in zones Z1B and
without delay when a feeder is closed onto a Z4 (see Fig. 7).
high-current short-circuit. It serves primarily as
fast-acting protection when connecting an n 5. Summary:
earthed feeder, but can also become effective
By division into two distance protection zones
(settable) with every closing – including auto-
(Z1B and Z4), selection of the cable and overhead
reclosure. The connection of the line is reported
line sections for double-end feeding in the event
to the protection by the “detection of the circuit-
of a fault is substantially simplified. In the practi-
breaker position” (parameter 1134).
cal application the auto-reclosure function can
In order to make use of the instantaneous high- only be performed restricted to the overhead line.
current tripping, the function must have been en- A fault in the cable section leads immediately to a
abled in the relay scope configuration. The value final TRIP command.
of the short-circuit current which leads to pickup
As shown, special requirements (such as selection
of the instantaneous tripping function is set as
of the faulty line section) can be implemented
‘I>>>’ value (parameter 2404). The value must be
easily and at low cost with the CFC logic in the
high enough to avoid the protection tripping
SIPROTEC distance protection.
(whatever the circumstances) in the event of a line
overload or current increase – e.g. as a result of a
brief interruption on a parallel line.
>AR block
(FNo. 2703)
LSP2690en.tif
Fig. 7 CFC plan for protection relay B


Redundant Supply with

Bus Coupler
Automatic switchover of incoming supply with

SIPROTEC 7SJ62
n 1. Introduction
Liberalized energy markets are demanding new so-
lutions for the operation of electrical systems. This
publication describes an application in which the
availability of power supply of a switchgear or plant
can be improved considerably by switching over
from a faulty to a redundant incoming supply. The
LSP2587.tif
influence of external system faults is minimized de-
cisively by fast disconnection of faulty system parts Fig. 1 SIVACON 400 V, with 7SJ62-protected and controlled circuit-breakers
and switchover from a faulty to a trouble-free in-
feed. These automation tasks can be accomplished
today with modern SIPROTEC protection relays The reliability of an electrical system is deter-
without the need for further equipment. mined by a number of factors. These include the
reliability of every single item of equipment, the
n 2. Influential variables of system availability kind and method of connection of the equipment,
“Power Quality” covers all the properties of an i.e. the system topology, the properties of the pro-
electrical power supply. Power quality can be fur- tection relays, the remote control equipment, the
ther subdivided into “voltage quality and system dimensioning of the equipment, the method of
reliability” as shown in Fig. 2. The latter is closely operation including troubleshooting, and the sys-
linked with an “adequate” power supply and the tem load capacity. The most frequently applied
security of the supply. Only the system reliability qualitative criterion in power system planning is
is looked at in detail below. (n-1), with which a system can be checked for
sufficient redundancy.
It requires that the system must be able to survive
Power failure of any item of equipment without imper-
Quality
missible restriction of its function. The (n-1) cri-
terion is a pragmatic and easy-to-handle basis for
decision but has the disadvantage that the supply
reliability cannot be quantified. Frequency, dura-
System Voltage
tion and scope of interruptions in the supply are
Reliability Quality not measured, with the result that it is not possible
(for example) to distinguish between different
(n-1)-safe system variants in terms of reliability.
Quantitative methods of system reliability analysis
Fig. 2 Subdivision of Power Quality
allow further evaluation of planning and operat-
ing variants supplementary to the qualitative
methods. The supply quality is quantified by suit-
able parameters and thus enables a comparative
assessment of different (n-1) reliable planning and
operation variants (for example). This allows a
specific estimate of the costs and benefits of indi-
vidual solutions in system planning and opera-
tion.

Switchover with redundant incoming feeders

means investment. However, by considering the
behavior in case of outages of equipment, the sys-
tem topology, the protection concepts, the system
capacity utilization (supply and loads) and the
method of operation, even more reliable and safe
system operation can be ensured. The aim is over-
all system reliability, expressed in terms of a high
degree of supply availability for special customers
with sensitive processes. Closer analyses by way of
the load cycle of individual feeders or transformer
stations – as well as permanent rationalization
measures in operation of the power systems – also
call for a higher degree of automation in all power Fig. 4 Typical failure costs per voltage sags
system sections.
2.1 Transient voltage sags and outages n 3. Functional principle and aim of automatic
The most frequent cause of system faults and in- switchover
ternal voltage sags or outages (total failures) is a A traditional method for a utility to solve its
lightning strike. As Fig. 3 shows, the system fault power quality problems is information from cus-
may be in the transmission system or in the distri- tomers about supply limitations suffered. With
bution system. the multifunction protection relays presented be-
Usually there is no total blackout but the remain- low it is possible to find solutions for protecting
ing residual voltage is greater than 70 %. whole areas from outages by means of protection
relays with integrated automatic functions.
Automatic switchover is suitable for disconnec-
ting an endangered supply and quickly bringing in
a redundant, secure supply with the aid of an al-
ternative incoming feeder. A fault is detected by
an undervoltage detection function. Using a direc-
tional overcurrent detection function, it can be
decided whether the fault is external or internal.
In the event of an external fault, switchover to the
alternative incoming feeder takes place. However,
if the fault is internal there is no switchover, with
the intention that the fault can be cleared by avail-
able circuit-breakers.
Switchover to the alternative incoming feeder or
the coupling of separated networks only takes
Fig. 3 Possible locations of system faults place instantaneously when both separated net-
works are synchronized. Otherwise it waits until
The economic damage caused by sags or outages is synchronicity between the two separated net-
immense (Fig. 4). The following Fig. 5 shows works is established or the voltage has dropped to
computer loads can fail already when the system such an extent that safe connection is possible.
amplitude deviates from its rating for less than However this is only on condition that the two in-
one period. This so-called ITI/CBEMA curve is coming feeders are not impaired in their voltage
used worldwide as a reference for the sensitivity of quality by the same system fault, such that
other load types too, because the appropriate switchover provides no protection against load
manufacturer data are often unavailable. The dif- shedding.
ficulty in protecting a highly automated factory is
largely attributable to the large number of loads
and the degree of networking of these loads.

In the case of rapid system decoupling (opening

the circuit-breaker in the faulty incoming feeder)
it can be assumed that the fault current has a
higher displacement than in normal switching.
This should be taken into account in the choice of
circuit-breakers. The system configuration and
the specific requirements regarding switching time
must therefore be analyzed before choosing the
suitable rapid switchover.
The following typical applications are particularly
conceivable:
1. Switchover from one redundant incoming
feeder to the next to protect loads from volt-
age outages
2. System decoupling in the event of a fault on
the load side and therefore prevention of the
fault from affecting other loads.
3.1 Practical principle

The SIPROTEC relays attend to full protection of
the incoming feeders by means of directional
overcurrent-time protection.
Configuration instructions for protection of the
incoming feeders are not dealt with in detail here.
Automatic switchover is implemented by at least
two autarchic SIPROTEC 4 relays (e.g. 7SJ62)
which can be adapted individually to the design
and basic conditions applying in a customer speci-
fication, in combination with the existing
Fig. 5 ITI/CBEMA curve
switchgear.
The following switchover possibilities can be
distinguished here:
n Overlap switchover
Both circuit-breakers are actuated almost
simultaneously
n Rapid switchover
Circuit-breaker 1 is opened and circuit-breaker 2
closed as long as the voltage is below ∆U – motor
rundown behavior is taken into account
n Slow switchover
Motors must have run down, or else be switched
back on as from a certain residual voltage, the Fig. 6 Bay layout
reason being high start-up current of the motor
groups; this possibility should be rare.
3.2 Description The relays are interconnected by binary signal

The desired configuration can be selected as “nor - communication between the binary inputs and
mal operation” with the preselector switch S100. outputs. In this way every relay can communicate
The selected circuit-breaker remains defined as with the other two relays and exchange informa-
“normally OPEN”. This open circuit-breaker is tion about circuit-breaker position and protection
considered as a backup in the event of a fault functions.
which can then supply the faulty, disconnected
busbar section with energy again. Each circuit-
breaker operates autarchically and is controlled by
one single multifunction relay.

Therefore it is possible to create self-controlling This configuration was installed in the plant of a
automatism also allowing manual control from petrochemical industry customer and has been
the outside. When connecting, the synchro-check operating reliably since 2002.
can be performed by the multifunction relays
The principle has proven so reliable that it is used
(7SJ64) themselves or by a separate synchro-check
in all the busbars there, from the 400 V switchgear
device.
through 6.6 kV right up to the 33 kV level.
a) In the event of undervoltage and breaker
failure from subordinate feeders or from the
parallel incoming supply, the circuit-breakers
are tripped individually by any protective
pickup.
b) If the protection has picked up due to a fault
outside the switchgear or plant, or the supply
voltage drops although there is no short-
circuit/earth fault, the parallel incoming
feeder is granted release (release of infeed B)
to close.
c) If the disconnector is closed in the trouble-
free incoming feeder and the parallel supply is
LSP2586.tif
released (2 releases), the circuit-breaker is
closed, either at synchronicity or if there is no
voltage on the busbar. Disconnection of the Fig. 8 Air-insulated switchgear 8BK, 6.6 kV, with
7SJ63-protected and controlled circuit-breakers
faulty incoming feeder and switch-in of the
substitute incoming feeder can be coordi-
nated by the timer T1 (overlap time). n 4. Summary
Multifunction relays which also assume control
By setting the post-fault time with timer T2, and protection duties for the switchgear or plant
the maximum permissible time interval is are highly attractive due to their greater flexibility.
specified which may pass between connection There is considerable interest in solutions for pro-
and the last satisfied synchronization condi- tection against outages that would otherwise bring
tion. whole factories to a standstill. Therefore this solu-
tion has the potential for use in both the low and
the medium-voltage sector.
Automatic switchover based exclusively on
SIPROTEC 4 relays represents an attractive alter-
native to existing products in terms of both invest-
ment volume and engineering effort. The neces-
sary functions are available. The integrated logic
can be used to great advantage for the parameteri-
zation (by means of a CFC logic editor) of auto-
matic switchover in the relays.
Fig. 7 Logic example for input field A
d) If the circuit-breaker in the faulty incoming

feeder has not opened properly, the circuit-
breaker in the substitute incoming feeder will
be reclosed by the breaker failure protection.

Coordination of Inverse-
Time Overcurrent Relays
with Fuses
n 1. Introduction
The duty of protection equipment is to allow
overload currents that occur during operation,
yet to prevent impermissible loading of lines and
equipment. To avoid damages in the case of short-
circuits the relevant equipment must be tripped in
the shortest possible time. On the other hand only
as few feeders or loads as possible should be dis-
connected from supply.
The protection relays available in the power sys-
tem must recognize the fault, perform tripping
themselves or give trip commands for the relevant
switching device.
The protection relays must be set to ensure selec-
tive tripping. Absolute selectivity is not always as-
sured. “Selectivity” means that the series-connec-
ted protection relay nearest the fault first trips the
faulted line. Other protection relays (further up- Fig. 1 Block diagram
stream) recognize the fault but trip only after a de-
lay (backup protection). However, they are not used as overload protection
In the following the use of HV HRC fuses because they can only trip reliably as from their
(high-voltage-high-rupturing capacity) and in- minimum breaking current. For most HV HRC
verse-time overcurrent-time protection relays fuse links the lowest breaking current is
(as well as their interaction) will be described. Imin = 2.5 to 3 x IN.
See Fig. 1. With currents between IN and Imin HV HRC fuses
cannot operate.
n 2. Protective equipment
2.1 HV HRC fuses When choosing HV HRC fuse-links, stressing of
the fuse from earth-fault current or residual cur-
The high-voltage-high-rupturing capacity fuse is a
rent must be considered.
protective device suited for non-recurring shut-
down in medium-voltage switchgear, in which the HV HRC fuse-links are installed with high-voltage
current is interrupted by the melting of a fusible fuse-bases in the switchgear. They can also be in-
element embedded in sand. stalled in the built-on units of the switch discon-
nectors provided. By combining switch discon-
HV HRC fuses are used for short-circuit protec-
nector and HV HRC fuse, the IN to Imin current
tion in medium-voltage switchgear up to 20 kV.
which is critical to the fuse can also be reliably
Used upstream of transformers, capacitors and
broken. The switch is tripped by the fuse's striker
cable feeders, they protect equipment and system
and disconnects the overload current in the three
components from the dynamic and thermal ef-
phases. Some typical breaking characteristics are
fects of high short-circuit currents by shutting
shown in Fig. 2.
them down as they arise.

2.2 Inverse-time overcurrent protection

Overcurrent protection is the main function of the
7SJ6 product range. It can be activated separately
for phase and earth-fault currents.
The I>> high-set overcurrent stage and the I>
overcurrent stage always work with definite trip-
ping time.
In the Ip inverse-time overcurrent stage, the trip-
ping time depends on the magnitude of the
short-circuit current.
Fig. 3 shows the basic characteristics of definite
and inverse-time overcurrent protection.
For inverse-time overcurrent functions (Ip stages)
various tripping characteristics can be set.
n Normal inverse (NI)
n Very inverse (VI)
n Extremely inverse (EI)
n Long time inverse (LI)
All characteristics are described by the formulae
below. At the same time, there are also distinc-
tions as follows:
IEC/BS ANSI
NI
t=
0.14
⋅ Tp  8.9341 
Breaking t= + 0.17966  ⋅ D
characteristics ( I / I p )0 . 02 − 1  (I / Ip )
2 . 0938
−1 
of HV HRC fuses
VI
t=
13.5
⋅ Tp  3.922 
t= + 0.0982 ) ⋅ D
(I / Ip ) − 1  (I / Ip ) − 1
2

EI
t=
80
⋅ Tp  5.64 
t= + 0.02434  ⋅ D
( I / I p )2 − 1  (I / Ip ) − 1
2

LI
t=
120
⋅ Tp  5.6143 
t= + 2.18592  ⋅ D
(I / Ip ) − 1  (I / Ip ) − 1 
t = Tripping time
Tp = Setting value of the time multiplier
I = Fault current
Ip = Setting value of the current
Table 1 IEC/BS and ANSI
The general IEC/BS characteristic is shown in

Fig. 4 and that of ANSI in Fig. 5
Fig. 3
Definite and
inverse-time
characteristics

Fig. 4 IEC/BS characteristics Fig. 5 ANSI characteristics
n 3. Network circuit and protection concept n 4. Selection and setting of protective components
The topology of a distribution system should be as The HV HRC fuses are selected using tables that
simple and clear as possible and ensure a reliable take into account transformer power (Sn), short-
supply. circuit voltage (Usc) and rated voltage on the HV side.
Using the short-circuit currents detected, a proposal
Individual transformer stations are supplied by
can be worked out for selective protection setting of
ring cables. An example of a ring cable system is
the inverse-time overcurrent functions:
shown in Fig. 6.
In order that a fault does not cause the whole ring n Ip must be set above the permissible rated current of
with all stations to fail, an “open” operating the cable (around 1.5 x IN cable)
method is the standard. In this example, trans- n I>> should not trip in the case of a fault on the low-
formers are protected on the low voltage (LV) voltage side
side with HV HRC fuses and the ring cable itself n In the case of a max. short-circuit current in the MV
with an overcurrent-time relay. system, there must be an interval of at least 100 ms
between the tripping characteristic of the HV HRC
3.1 Calculating the relevant system currents fuse and the inverse-time characteristic. The time
The full load current and short-circuit strength are multiplier Tp must be set to get this safe grading
the selection criteria for the cable to be used. The time.
transformer rated currents must not deviate too It must be borne in mind that the value of the time
much from the rated currents of the cables used. multiplier Tp (in 7SJ6 from 0.05 to 3.2 seconds) does
The maximum and minimum short-circuit cur- not correspond to the genuine tripping time of the
rents (3, 2, 1 phase) appearing in this power sys- characteristic. Rather, the inverse-time characteristic
tem section must be calculated before the para- can be shifted in parallel in the time axis by this value.
meters of the relays can be set. LV-side short-
circuit currents must also be taken into account
here. It is advisable to use programs such as
SIGRADE (Siemens Grading Program) to calcu-
late the short-circuit currents.
For further information please visit us at:
www.siemens.com/systemplanning

Fig. 6 Example of a 20 kV ring distribution system
n 5. Proof of selective tripping current bands always end on the voltage current
As mentioned earlier, selectivity means only the scale. The right-hand characteristic in a band is
protection relay closest to the faulty system sec- the maximum short-circuit current (3 phase), cal-
tion trips. Protection equipment connected (up- culated (here in green) from the incoming ele-
stream) in series must register the fault but only ments (generators, transformers, etc). The left-
trip after a delay period. Typically, proof of selec- hand characteristic shows the minimum short-cir-
tive tripping is shown in a current-time-diagram cuit current (1 or 2 phase) which is calculated on
with double logarithm scale. Programs like the basis of the impedances of the elements in the
SIGRADE are also used for this. power system up to the location of the fault.
For the power system sections in question, typical Band 1 (Transf. D Pr) shows the bandwidth of the
or critical time grading diagrams are selected. 20 kV power system;
Each protection relay has its own name, which de- band 2 (Transf. D Sec) shows that of the 0.4 kV
scribes the installation location. The same power power system.
system and protective elements shown in more The above-mentioned bands are contained in the
than one time grading diagram have the same time-grading diagrams (Figs. 7 to 11).
name.
n 6. Grading of overcurrent-time relay and
The color of the name in the time grading path HV HRC fuse
(left side of the diagram) matches the color of the
As an example of the power system shown in
set characteristic (in the time grading diagram on
Fig. 6, in 3 time sequence diagrams the most usual
the right) in the current-time-diagram. On the left
characteristics (NI, VI, EI) of the inverse-time
side, in addition to the single-line circuit diagram
overcurrent protection are shown with the corre-
(time-grading path) for each protection relay, the
sponding HV HRC fuses characteristic. The over-
type name, the setting range and the set values are
current-time relay 1, HV HRC fuse D and trans-
given.
former D are selected from the circuit diagram.
In addition to the characteristics of the protection
relay, the current-time-diagram shows the short-
circuit current ranges plotted with minimum and
maximum values as bandwidth (values from the
short-circuit calculation). These short-circuit

Setting range Setting
Ip = 0.1 – 4 A Ip = 1.8 A
Tp = 0.05 – 3.2 s Tp = 0.05 s
Ip_mi = 2.3 kA Ip_ma = 7.8 kA

Is_mi = 15.8 kA Is_ma = 18.2 kA
Overcurrent-time relay with “Normal Inverse” (NI) setting
Band 2 Band 1
Fig. 7 Time-grading
diagram, inverse-time NI

Ip = 0.1 – 4 A Ip = 1.8 A
Tp = 0.05 – 3.2 s Tp = 0.15 s

Is_mi = 15.8 kA Is_ma = 18.2 kA
Overcurrent-time relay with “Very Inverse” (VI) setting
Band 2 Band 1
Fig. 8 Time-grading
diagram, inverse-time VI

Ip = 0.1 – 4 A Ip = 1.8 A
Tp = 0.05 – 3.2 s Tp = 0.45 s

Is_mi = 15.8 kA Is_ma = 18.2 kA
Overcurrent-time relay with “Extremely Inverse” (EI) setting
Band 2 Band 1
Fig. 9 Time-grading
diagram, inverse-time EI

Ip = 0.1 – 4 A Ip = 1.8 A
Tp = 0.05 – 3.2 s Tp = 0.05 s

Is_mi = 15.8 kA Is_ma = 18.2 kA
Overcurrent-time relay with VI setting,

with same setting as NI
Band 2 Band 1
Fig. 10
Time-grading diagram,
very inverse, with
setting like normal
inverse

Ip = 0.1 – 4 A Ip = 1.8 A
Tp = 0.05 – 3.2 s Tp = 0.05 s

Is_mi = 15.8 kA Is_ma = 18.2 kA
Overcurrent-time relay with EI setting,

with same setting as NI
Band 2 Band 1
Fig. 11
Time-grading diagram,
extremely inverse, with
setting like normal in-
verse

For the transformer considered (20/0.4 kV, n Conclusion

Sn = 0.8 MVA, Uk = 6.1 %), a 63 A HV HRC fuse The steeper the slope of the characteristic the
is selected according to the above-mentioned se- lower the tripping time with maximum fault
lection tables. current. The safe grading time from the HV
In order to maintain selectivity, the target trip HRC characteristic becomes smaller. The coor-
time is approx. 100 ms for the overcurrent-time dination shown here of the protection devices is
relay setting in the various characteristics with only part of the power system and must be
maximum short-circuit current on the 20 kV side. adapted to the concept of the overall power sys-
Under the same short-circuit conditions the tem with all protection relays.
HV HRC fuse trips in approx. 1 ms.
Note:
By setting the Ip appropriately, the maximum fault In this example there was no setting of I>>, be-
on the LV side does not lead to pickup of the over- cause the inverse-time characteristics themselves
current-time relay. As can be seen in Fig. 10 the trip in the ≤ 0.2 s range in the event of the maxi-
characteristic begins on the right, next to the max- mum or minimum 20 kV side fault.
imum short-circuit current (brown, vertical lines).
The following setting values should be used to n 7. Summary
achieve a safe grading time of 100 ms between all This application example demonstrates the engi-
types of o/c inverse-time characteristics (NI, VI, neering effort necessary to achieve a selective time
EI) and the characteristic of the fuse. grading.
Real power systems are more complex and equip-
Fig. Ip x IN Tp (s) Characteristics ped with various protection relays. Whatever the
7 1.8 0.05 Normal inverse (NI) circumstances, it is necessary to know the operat-
8 1.8 0.15 Very inverse (VI)
ing mode of the power system (parallel, generator,
meshing, spur lines etc) as well as to calculate the
9 1.8 0.45 Extremely inverse (EI) rated and short-circuit currents. It is worth the ef-
Table 2 fort for the protection engineer because the objec-
tive is to lose only the faulty part of the power sys-
When comparing the three figures it is clear that tem.
the area between the HV HRC and inverse-time
characteristics is smallest for the setting NI. SIGRADE software effectively supports grading
Therefore the NI setting must be preferred in this calculations. Power system planning and time
example. grading calculation is also offered by Siemens.
In order to explain the difference in the character-
istics more clearly, two diagrams are shown with
n 8.References
Günther Seip: Elektrische Installationstechnik
the characteristics VI and EI with the same setting
values as NI. Siemens: Manual for Totally Integrated Power
Catalog HG12: HV HRC Fuses
Very inverse (VI) Extremely inverse (EI)
Ip = 1.8 Ip = 1.8 SIGRADE Software V3.2
Tp = 0.05 Tp = 0.05 Manual 7SJ61: Multifunctional Protective Relay
with Bay Controller 7SJ61
See Fig. 10 See Fig. 11
Table 3

Medium-Voltage Protection
with Auto-Reclosure and
Control
n 1. Introduction
An important protection criterion in medium-
voltage applications is overcurrent-time protec-
tion. Hardware redundancy can be dispensed with
in favor of lower-cost solutions, thanks to numer-
ical technology and the high reliability of the
SIPROTEC 4 protection relays. The SIPROTEC 4
protection relays also allow functions which go
beyond the basic scope of protection:
n Unbalanced load (negative-sequence) pro-
tection, motor protection functions, circuit-
breaker failure protection,...
n Other voltage-dependent protection functions
such as voltage protection, directional
overcurrent protection
n Auto-reclosure
R_HA25-328.tif
n Control, including interlocking
n Integration in a control system
This enables all the requirements in the feeder to
be met with a single relay. Scalable, flexible hard- Fig. 1 SIPROTEC medium-voltage protection
ware allows simple adaptation to any application.
n 2. Protection concept
2.1 Overcurrent-time protection
The task of overcurrent-time protection is to detect
the feeder currents, in order to initiate tripping by
the circuit-breaker in the event of overcurrent. Se-
lectivity is achieved here by current grading or time
grading. The phase currents IL1, IL2 and IL3 and the
earth current IE serve as measuring variables here.
(Non-directional) overcurrent-time protection is
used in medium-voltage power systems with
single-end infeed or as backup protection in
high-voltage applications.
Fig. 2 Block diagram

2.1.1 Grading and selectivity 2.1.3 Inverse-time overcurrent protection

The aim of every protective setting is to achieve The inverse-time characteristic is widely used in
selectivity, i.e. the protection relay closest to the countries with a British and American influence.
fault location trips the CB; all the others detect the Here the delay time is dependent on the current
fault but do not switch off or at least only after a detected.
delay. This ensures backup protection if “regular”
Inverse-time overcurrent protection characteristics
protection fails.
according to IEC 60255
There are basically two criteria for achieving selec- IEC 60255-3 defines four characteristics which
tivity: differ in their slope.
n Time Inverse
Here a protection relay initiates tripping imme- Very Inverse
diately or with an adjustable delay time. Since Extremely Inverse
the power system fault is usually detected by a Long Inverse
number of protection relays in the power
system, the protection relay with the shortest The calculation formulae and the corresponding
delay time initiates tripping. The delay times in characteristics are shown below by way of com-
the individual protection relays are defined such parison.
that the short-circuit is cleared by the protec- .
014
Inverse t= ⋅ Tp
(I I p ) − 1
0 . 02
tion relay closest to the fault.
This type of grading is normally used for cable and 135
.
Very inverse t= ⋅T
overhead power line systems.
( p ) −1 p
I I
n Current
Another grading criterion may be the magni- 80
Extremely inverse t= ⋅ Tp
(I I )
2
tude of the short-circuit current itself. Since the p −1
size of the short-circuit current cannot be deter-
mined exactly in pure line or cable systems, this 120
Long inverse t= ⋅T
method is used for grading of transformers. The ( Ip ) − 1 p
I
transformer limits the short-circuit current re-
sulting in different magnitudes of short-circuit
current on the high and low-voltage side.
This behavior is utilized to achieve selectivity in
tripping, as is attained in time grading.
The flexibility of SIPROTEC 4 overcurrent-time
protection relays allows a mixture of these two cri-
teria and therefore helps to achieve optimum sup-
ply security.
2.1.2 Definite-time overcurrent protection

Fig. 2 Definite-time overcurrent protection is the main
2-stage definite-time characteristic used in Europe (except in countries
overcurrent characteristic with a British influence). Delay times are assigned
(I>>, I>)
to several current pickup thresholds.
I>> pickup value (high short-circuit current)
t>> (short delay time)
I> pickup value (low short-circuit current),
t> (delay time)
Fig. 3 Comparison of inverse-time overcurrent

protection characteristics

The corresponding characteristic is selected de- Normally only one single three-pole reclosure is
pendent on the overall grading coordination performed for medium-voltage applications. Dead
chart. However, the inverse characteristic is suffi- times between 0.3 and 0.6 s usually suffice for ade-
cient for most applications. quate de-ionization of the flashover distance and
thus a successful reclosure.
Inverse-time overcurrent protection characteristics
according to ANSI/IEEE The lockout times (time up to next reclosure) are
Characteristics are also defined by ANSI/IEEE chosen so that protection relays affected by the
similar to those according to IEC 60255. For power system fault have reliably reset. In the past
further details about these, see the application this led to relatively long lockout times (approxi-
example on “Coordination of Inverse-Time mately 30 s) due to the dropout time of mechani-
Overcurrent Relays with Fuses”. The ANSI char - cal protection relays. This is not necessary in nu-
acteristics are also available as standard in all merical protection relays. Shorter lockout times
SIPROTEC 4 overcurrent-time protection relays. can therefore reduce the number of final discon-
nections (unsuccessful reclosures), for example
2.1.4 User-defined characteristics during thunderstorms.
Numerical protection relays like SIPROTEC 4 also In the past separate relays were used for
allow the user to freely define characteristics, and protection and automatic reclosure. The initiation
therefore enable maximum flexibility. This means for this was given by parallel wiring with the pro-
ease of adaptation to existing protection concepts, tection relay. In SIPROTEC 4 relays the auto-
e.g. when renewing the protection, even for special reclosure function can be integrated in the protec-
applications. tion relay; there is no need for any additional relay
and wiring.
2.1.5 Combined characteristics
SIPROTEC 4 overcurrent-time protection allows 2.3 Control
the advantages of definite and inverse-time There is a noticeable worldwide trend towards
overcurrent protection to be combined. On the automation, even in medium-voltage power sys-
one hand, with the high-set current stage I>>, the tems. SIPROTEC 4 protection relays provide the
tripping time with high short-circuit currents can conditions for controlling the feeder both locally
be reduced in comparison with inverse-time and remotely by telecontrol/station control and
overcurrent protection characteristics, and on the protection systems. This is supported by the ap-
other hand the grading can be adapted optimally propriate control elements on the relay and vari-
to the characteristic of the HV HRC fuses with in- ous serial interfaces. See Chapter 4 for further
verse-time overcurrent protection characteristic. information.
2.1.6 Sensitivity n 3. Settings

The earth current can be measured or calculated The determining of the most important setting
in addition to the phase currents. Independent parameters is explained in this chapter by means
protection stages for phase-to-earth faults are also of a typical application.
available in the SIPROTEC 4. As a result, sensitiv-
ity below the rated current is achieved for such a 3.1 Overcurrent-time protection
fault. The setting of the overcurrent stages is defined by
the grading coordination chart of the overall net-
2.2 Auto-reclosure work. Current grading is possible for the “trans -
Auto-reclosure is only used on overhead lines, be- former” protection object; only time grading can
cause the chances of success are relatively slight in usually be applied for overhead lines/cables.
the event of faults in a cable network. About 85%
of reclosures are successful on overhead lines, 3.1.1 High-set current stage I>>
which contributes greatly to a reduction in power The high-set current stage I>> is set under the ad-
system downtimes. dress 1202 and the corresponding delay T I>> un-
Important parameters for reclosure are: der 1203. It is normally used for current grading at
high impedances such as are encountered in trans-
Dead time formers, motors or generators. Setting is such that
Lockout (blocking) time it picks up for short-circuits reaching into this im-
Single or three-pole pedance range.
Single or multishot

Example:
Transformer in the infeed of a busbar with the fol-
lowing data:
Rated apparent power SNT = 4 MVA
Short-circuit voltage Uk = 10 %
Primary rated voltage UN1 = 33 kV
Secondary rated voltage UN2 = 11 kV
Vector group Dy 5
Neutral earthed
Short-circuit power
on 33 kV side 250 MVA
LSP2682.tif
The following short-circuit currents can be calcu-
lated from these data:
3-pole, high-voltage side short-circuit Fig. 4 DIGSI parameter sheet, definite-time overcurrent protection
I"SC3 = 4389 A phase
3-pole, low-voltage side short-circuit

I"SC3, 11 = 2100 A 3.1.2 Overcurrent stage I>
on the high-voltage side flow The maximum operating current occurring is sig-
I"SC3,33 = 700 A nificant for the setting of the overcurrent stage I>.
Pickup by overload must be ruled out because the
rated current of the transformer HV
relay works as a short-circuit protection with cor-
INT, 33 = 70 A (high-voltage side)
respondingly short operating times in this mode
rated current of the transformer LV and not as overload protection. It is therefore set
INT, 11 = 211 A (low-voltage side) for lines at about 20 % and for transformers and
current transformer (high-voltage side) motors at about 40 % above the maximum (over)
INW, 33 = 100 A / 1 A load to be expected. The delay to be set (parame-
current transformer (low-voltage side) ter 1205 TI>) is given by the grading coordination
INW, 11 = 300 A / 1 A chart created for the power system.
Due to the setting value of the high-set current The set time is a purely additional time delay
stage I>> which does not include the operating time (mea-
I>>/IN >
1 I
⋅ N Transfo suring time). The delay can be set to ∞. The stage
U k Transfo I N CT then does not trip after pickup but the pickup is
signaled. If the I> stage is not needed at all, the
the following setting on the protection relay: pickup threshold I> is set to ∞. Then there is
The high-set current stage I> must be set higher neither pickup indication nor tripping.
than the maximum short-circuit current detected According to the above example this gives a calcu-
on the high-voltage side in the event of a fault on lated setting value of
the low-voltage side. In order to attain a sufficient
noise ratio even at fluctuating short-circuit power, I > = 14
. ⋅ I NT, 33 = 14
. ⋅ 70 A =100 A =1.0 ⋅ I NW, 33
a setting of
3.1.3 Inverse-time overcurrent protection stages Ip
I>>/IN = 10, i.e. I>> = 1000 A It must be taken into account when selecting an
is selected. inverse-time overcurrent characteristic that a fac-
tor of approximately 1.1 is already incorporated
Increased inrush current surges are disarmed by between the pickup value and the setting value.
the delay times (parameter 1203 T I>>) provided This means that pickup only takes place when a
their fundamental component exceeds the setting current 1.1 times the setting value flows. The cur-
value. The set time is a purely additional time de- rent value is set under address 1207 Ip. The maxi-
lay which does not include the operating time. mum operating current occurring is significant
for the setting.

Pickup by overload must be ruled out because the

relay works as a short-circuit protection with ap-
propriately short command times in this mode
and not as overload protection. The correspond-
ing time multiplier is accessible under address
1208 T Ip (51 TIME DIAL) when selecting an IEC
characteristic, and under address 1209 51 TIME
DIAL when selecting an ANSI characteristic. This
must be coordinated with the grading coordina-
tion chart of the power system.
The time multiplier can be set to ∞. Then the stage
does not trip after pickup but the pickup is sig-
naled. If the Ip stage is not needed at all, address
LSP2683.tif
1201 DMT/IDMT PHASE = DTM only (FCT
50/51) is selected in the configuration of the pro-
tection functions. Fig. 5 DIGSI parameter sheet, inverse time overcurrent protection phase
3.1.4 Earth current stages

IE>> (earth)
The high-set current stage IE>> is set under ad-
dress 1302 (50 N-2 PICKUP) and the correspond-
ing delay T IE>> under 1030 (50 N-2 DELAY).
Similar considerations apply for the setting, as
previously described, for the phase currents.
IE> (earth) or IEp
The minimum occuring earth fault current is
mainly decisive for the setting of the overcurrent
stage IE> or IEp. If great inrush currents are to be
expected when using the protection relay on
transformers or motors, an inrush restraint can be
LSP2684.tif
used in 7SJ62/63/64 for the overcurrent stage IE>
or IEp. This is switched on or off for both phase
and earth current together under address 2201
DIGSI parameter sheet, definite time overcurrent earth protection
INRUSH REST.
The time delay to be set (parameter 1305 T IE>/ A description for setting the most important
50 N-1 DELAY or 1308 T /IEp/51N TIME DIAL) is reclosure parameters follows:
given by the grading coordination chart created
for the power system, whereby a separate grading 7105 Time restraint:
coordination chart with shorter delay times is of- The blocking time TIME RESTRAINT (address
ten possible for earth currents in the earthed 7105) is the time span following successful rec-
power system. losure after which the power system fault is con-
sidered cleared. Generally, a few seconds are
3.2 Auto-reclosure enough. In regions with frequent thunderstorms a
The integrated auto-reclosure function can be short lockout time is recommendable, to reduce
used for performing reclosures on overhead lines. the danger of final disconnection due to lightning
This can be initiated by every overcurrent stage strikes in rapid succession or cable flashover. The
and other protection functions. External initiation default selection is 3 s.
via binary inputs is also possible. In this way the
reclosing function can be adapted individually to
the respective application without external wiring.

7117 Action time

The action time checks the time between the
pickup of a relay and the trip command of a pro-
tection function parameterized as a starter, in
ready (but not yet running) auto-reclosure.
If a trip command is received from a protection
function parameterized as a starter within the ac-
tion time, auto-reclosure is initiated. If this time is
outside the parameterized value of T-ACTION
(address 7117), auto-reclosure is blocked dynami-
cally. With inverse-time characteristics the release
time is determined essentially by the fault location
LSP2686.tif
and the fault resistance. With the help of the ac-
tion time, no reclosure is performed in the event
of very remote or high-resistance faults with a Fig. 8 DIGSI parameter sheet, auto-reclosure configuration
long tripping time.
Presetting of ∞ always initiates a reclosure. The dead time before the 1st reclosure for the
7135 Number of reclosure attempts, earth reclosure program “Phase” (phase-to-phase fault)
7136 Number of reclosure attempts, phase is set in address 7127 DEADTIME 1:PH; for the
The number of reclosures can be set separately for reclosure program “Earth” (single phase-to-earth
the programs “Phase” (address 7136, NUMBER RC fault) it is set in address 7128 DEADTIME 1:G.
PHASE/# OF RCL. PH) and “Earth” (address 7135 The duration of the dead time should relate to the
NUMBER RC EARTH/# OF RCL. GND). The pre- type of application. For longer lines the time
setting for both parameters is 1 (one); one reclosure should be long enough for the short-circuit arc to
cycle is therefore executed. extinguish and de-ionize the ambient air, to allow
successful reclosure (usually 0.9 s to 1.5 s). The
stability of the power system has priority in the
case of lines fed from several ends. Since the dis-
connected line cannot develop any synchronizing
forces, often only a short dead time is permissible.
Normal values are between 0.3 s and 0.6 s. Longer
dead periods are usually allowed in radial systems.
The default is 0.5 s.
LSP2561en.tif
DIGSI parameter sheet, auto-reclosure (general)
The “configuration sheet” defines which of the

protection stages starts the reclosure. For each of
LSP2687.tif
the stages it can be decided whether this stage

starts the reclosure, does not start it or blocks it
out. Fig. 9 DIGSI parameter sheet, auto-reclosure (1st reclosure cycle)
7127 Dead time 1: ph
7128 Dead time 1: G
The parameters 7127 and 7128 define the length
of the dead times of the 1st cycle. The time defined
by the parameter is started upon opening the
circuit-breaker (if auxiliary contacts are allocated)
or upon reset after the trip command.

n 4. Further functions
As already described in Chapter 2, a number of
additional functions can be configured in the
SIPROTEC 4 relays. Apart from further protec-
tion functions, these also include control tasks for
the feeder. All SIPROTEC 4 relays (e.g. 7SJ61 and
7SJ62) have 4 freely assignable function keys F1 to
F4 which simplify frequently required operations.
These function keys can take the user directly to
the display window for measured values, or to
fault event logs for example. If the relay is also to
be used for feeder control, these keys can be used
for controlling the circuit-breaker. The key F1
then selects the ON command for example, key F2
the OFF command and key F3 executes the se-
lected command (two-stage command output).
LSP2564.tif
Fig. 12 Example of interlock logic
The switching authority (local/remote) can be

LSP2299.eps
changed and the interlock check overridden by

two key-operated switches.
Fig. 10 Front view 7SJ61 or 7SJ62
The 7SJ63 also has a graphic display on which the

individual feeder mimic diagram can be shown.
Separate ON/OFF control buttons ensure safe and
reliable local control. An integrated, freely pro-
grammable interlock logic prevents switching
LSP2565.tif
errors.
Fig. 13 Front view, key-operated switch with

customer-specific feeder mimic diagram
LSP2316.eps
Fig. 11 Front view 7SJ63

n 5. Connection examples If voltages are also available (from the feeder or

5.1 Current and voltage transformers as a busbar measurement), these can be connected
Connection of the protection relays to the on 7SJ62/63/64 and then also enable voltage-
switchgear depends on the number of switching dependent protection functions (directional over-
objects (circuit-breakers, disconnectors) and cur- current protection, voltage protection, frequency
rent and voltage transformers. Normally at least protection, ...).
three current transformers are available per feeder
which are connected to the protection relay as fol-
lows.
Fig. 14 Transformer connection to three current trans-

formers
In some systems, also the earth current is mea-

sured by a core-balance current transformer. This
can be connected to the protection relay sepa-
rately. A core-balance current transformer achie-
ves greater accuracy (sensitivity) for low earth
currents. Fig. 16 Transformer connection to three current and three voltage
transformers
5.2 Input/output periphery

In addition to the current transformers (and if
required to the voltage transformers too), at least
the TRIP command has to be wired to the circuit-
breaker. The standard allocation supports this by
practice-oriented preassignment.
Preassignment of the inputs and outputs in the
7SJ610:
Binary inputs
BI1 Block definite/inverse time
overcurrent protection
BI2 LED reset
BI3 Display lighting on
Binary outputs (command relays)
BO1 TRIP command
BO2 Reclosure command
Important: The cable shield must be earthed on the cable side! BO3 Reclosure command
Note: Reversing the current polarity (address 0201) also reverses BO4 MV monitoring
the polarity of the current input IE
Fig. 15 Transformer connection to three current trans-

formers and core-balance CT

LEDs
LED1 TRIP command
LED2 PICKUP L1
LED3 PICKUP L2
LED4 PICKUP L3
LED5 PICKUP E
LED6 MV monitoring
LED7 Not used
The assignment can be changed and the pro-
tection parameters set conveniently with the
DIGSI 4 operating program. The parameteri-
zation data can then be saved and copied
conveniently as a basis for further feeders.
n 6. Summary
LSP2688.tif
SIPROTEC 4 protection relays are suitable for al-
most any application due to their modular hard-
Fig. 17 Configuration matrix
ware structure and the flexible scope of functions.
A suitable relay with the necessary scope can be
selected in line with requirements. Factory para-
meterization is oriented to typical applications
and can often be adopted with only small modifi-
cations. In the parameter setting with DIGSI, all
unnecessary parameters are hidden so that clarity
is much improved.
The retrofitting of serial interfaces for subsequent
integration into a substation control and protec-
tion system is also possible locally, which reduces
downtimes to a minimum. The functional scope
can also be changed later by “downloading” a new
order number.


Differential Protection of
Cables up to 12 km via
Pilot Wires
(Relay Type: 7SD600)
n 1. Introduction
Line differential protection systems make it possi-
ble to protect cables or overhead lines selectively
and as fast as possible in the event of a short-
circuit. The application domain of the SIPROTEC
7SD600 described here is predominantly in the
medium-voltage sector if either the tripping times
of graded overcurrent-time protection relays be-
come too great, or if distance protection relays are
LSP2001.eps
no longer able to guarantee the desired selectivity.
The 7SD600 digital differential protection relay
provides short-circuit protection for cables and
overhead lines in power supply systems, inde-
pendent of the system star-point configuration. It
works according to the conventional 2-conductors Fig. 1 SIPROTEC 7SD600 line differential protection relay
principle. Here, the phase currents at the two line-
ends are – with the help of summation current
transformers – added up to one summation cur-
rent. These are then transformed by voltage divid-
ers into proportional voltages, which are fed with
reversed polarity to two pilot wires. The resultant
voltage difference finally produces a current,
which represents the determinant tripping magni-
tude for both relays. Because of its rigorous local
selectivity (the protection range is limited by cur-
rent transformers at both ends of the line), differ-
ential protection is generally applied as an
instantaneous main protection since no other
protection can disconnect the line more quickly
and selectively.
2.1 Differential protection (ANSI 87L)

2.1.1 Principle and current transformer
connection
The differential protection function of the 7SD600
recognizes short-circuits in the protection range
by comparing the summation currents detected at Fig. 2 Principal circuit diagram of line differential
both ends of the line. In order to do this, the sec- protection 7SD600
ondary phase currents from the primary current
transformers are fed with variable weighting This also assumes, that current transformers with
(number of windings) into the summation cur- identical primary values are used at both ends,
rent transformer which combines them to pro- otherwise the variable windings ratio must be
duce a summation current. equalized by an appropriate arrangement of the
matching transformer and/or summation current
transformer.

2.1.2 Summation current transformers 2.1.3 Differential current

The 4AM4930 summation current transformer is With normal summation current transformer
installed as standard in the normal circuit. connection and symmetrical current flow (rated
quantity), 20 mA flow to the summation current
transformer on the secondary side. This summa-
tion current is now measured in the local 7SD600,
and also fed as voltage drop to two pilot wires via
an internal resistor of the protection relay. At the
remote end the summation current is formed in
the same manner and also fed as voltage drop to
both pilot wires, but this time with reversed pola-
rity.
In healthy state, the reverse polarity voltages
should neutralize each other out. However, under
fault condition there is in the case of different
summation currents a resulting voltage which will
drive a current (proportional to the theoretical
differential current) along both wires. This cur-
Fig. 3 4AM4930 summation current transformer normal rent is then measured by the protection relay and
connection
serves as a tripping parameter.
This summation current transformer has different
2.1.4 Transformers in the protected zone
primary windings with several tapping points al-
lowing phase current and connection type (for ex- The 7SD600 optionally also analyses the summa-
ample two-phase) mixing ratios to be varied. This tion currents for a component of 100 Hz. This
way, increased sensitivities (e.g. of the earth cur- makes it possible to expand the protection range
rent) or preferences in the case of double earth beyond a transformer. Additional external match-
fault can be neatly ensured. ing transformers must nevertheless ensure, that
the transformation ratio of the currents and their
Fault W W/ 3 I1 for IM = 20 mA phases is compensated analagously to the trans-
L1-L2-L3 (sym.) 3 1.0 1 x IN former.
L1-L2 2 1.15 0.87 x IN 2.1.5 Restraint current
L2-L3 1 0.58 1.73 x IN In order to stabilize the differential protection sys-
L3-L1 1 0.58 1.73 x IN tem against overfunction (unwanted operation) in
the event of external faults, the “differential cur -
L1-E 5 2.89 0.35 x IN
rent” tripping parameter is standardized to a sta -
L2-E 3 1.73 0.58 x IN bilization parameter. The latter is the sum of the
L3-E 4 2.31 0.43 x IN magnitudes of the currents detected at both ends
of the protection range. This stabilization means
Table 1 Fault types and winding valencies W in case
of normal connection L1 - L3 - E that in the event of large currents flowing through
and of measurement errors resulting from trans-
The selected summation current transformer con- former faults or transformer saturation, the trip-
nection must be implemented identically at both ping parameter must likewise be high.
ends of the line to avoid false tripping. Measuring the local summation current trans-
former secondary currents and the current flow-
ing through the pilot wires ensures, that each of
the two protection relays can calculate both the
differential and stabilizing currents and react ac-
cording to the tripping characteristic.
Fig. 4 Function diagram

2.1.6 Tripping characteristic 2.1.7 Pilot wires

The 7SD600 tripping characteristic consists of Symmetrical telecommunication wire pairs (typi-
three sections. In the area of small currents a fixed cally 73 Ω/km loop resistance and 60 nF/km ca-
tripping threshold settable as parameter must be pacity) with a wire/wire asymmetry (at 800 Hz) of
exceeded in order to ensure response. Both the less than 10-3 are suitable. The loop resistance may
other tripping characteristic branches have set de- not exceed 1200 Ω. Furthermore, the longitudinal
faults. Current-proportional transformer faults voltage component induced in the pilot wires by
rise as the current magnitude increases. This is the short-circuits (to earth) must be taken into ac-
taken into account in the tripping diagram with a count. The induced direct axis voltage component
section of a straight line through the origin with a can be calculated according to the following for-
1/3 gradient (slope). In the case of even greater cur- mula:
rents the tripping limit is determined by a further
U l = 2 π f ⋅ M ⋅ I k1 ⋅ l ⋅ r1 ⋅ r2
straight line, which intersects the stabilization axis
at 2.5 · INLine and has a gradient of 2/3. This branch where
Ul = Induced direct axis voltage component
takes account of incipient current transformer sat-
ƒ = Rated frequency [Hz]
uration. M = Mutual inductance between power cable and
pilot wires [mH/km]
Ik1 = Maximum single-pole short-circuit current [kA]
l = Length of the parallel distance between power
cable and pilot wires [km]
r1 = Reduction factor of the power cable
(in case of overhead lines r1 = 1)
r2 = Reduction factor of the pilot wire cable
The calculated induced voltage needs only be half

taken into account since it builds up on the insu-
lated pilot wires at both ends. If this exceeds 60 %
of the permitted test voltage, additional measures
(isolating transformers) are necessary. Isolation
(barrier) transformers are available for isolation
up to 5 kV and 20 kV respectively. The center tap
Fig. 5 Restraint characteristic of the differential pro-
tection drawn with a default of IDiff> = 1.0 · IN Line on the side facing the protection relay must be
earthed for anti-touch protection reasons, but the
IDiff = |I1 + I2| pilot wire connection must not be earthed or pro-
IRestraint = |I1| + |I2| vided with surge arresters.
I1 = Current at local line end, positive flowing into line
I2 = Current at remote end, positive flowing into line
IN Line = Line rated current
In order that differential protection remains stable

in the case of external faults with strong currents,
the 7SD600 offers additional stabilization with re-
gard to possible transformer saturation. Here, on
the basis of the stabilization and differential cur-
rent trend, the protection recognizes that an exter-
nal fault initially occurred before the build-up of
the differential current as a result of transformer Fig. 6 Connection of isolating transformers 7XR9515 (5 kV) or 7XR9513 (20 kV)
saturation. When the current values enter the ad-
ditional stabilization (restraint) range, the differ- 2.2 Backup protection functions (ANSI 50)
ential protection is blocked for a maximum of
As is usual with modern, numerical protection re-
1 second in order to give the transformer time to
lays, the 7SD600 also offers further integrated pro-
come out of saturation. However, if during this
tection and additional functions. The user must
time steady state prevails in the tripping range for
nevertheless be aware of the lack of hardware re-
two network periods, blocking is neutralized and
dundancy when deploying these functions. For
the protection makes the decision to trip.
this reason, at least a further separate short-circuit
protection relay, for instance a 7SJ602, should be
installed.

2.2.1 Overcurrent-time protection (ANSI 51) The 7SD600 is notable for its few setting parame-
The 7SD600 includes overcurrent-time protection ters, which allow it to be configured quickly and
alongside differential protection as an emergency easily. Pilot-wire monitoring is the only function
function, i.e. for cases where the main function is that can be activated (or deactivated) under
no longer available. Parameterization makes it “Scope of the device”, provided the relay has been
possible to set whether the emergency definite- ordered with this option. This must be activated.
time overcurrent protection should generally be
activated when differential protection is ineffec- 3.1 System /line data
tive, or only if the wire monitoring responds. This The parameters defined by the primary equipment
emergency definite-time overcurrent protection are set under the “system/line data” heading (see
works with the local summation current and fea- Fig. 7). These include network frequency, current
tures one single stage. The current threshold is set transformer ratio and minimum circuit-breaker
above the maximum symmetrical load current. activation time in the event of protective tripping.
Since in general no time grading is possible if full In order to better match the differential protec-
selectivity is to be retained, a compromise between tion characteristic, the protection parameters are
selectivity and speed of protection has to be referred to the line rated current; this must be in-
found. In any case the tripping time should be de- put at this point and must imperatively be the
layed by at least one grading, in order to wait to same in both relays. As already described above,
see whether this high current is caused by faults the resistance of the pilot-wire connection is re-
on adjoining power system sections and other quired for correct calculation of the current value
protection relays selectively trip upon this fault. at the remote end. This can be calculated either
from the pilot-wire connection data sheets or can
2.2.2 Additional functions be measured within the context of commissioning
Pilot-wire monitoring by the relay itself in accordance with the instruc-
The ohmic resistance of the pilot-wire loop is tions in the manual. This value must subsequently
needed for correct calculation of the summation be entered here. Finally, the lock-out function can
current at the remote end of the protection range. be switched either on or off at this point. The acti-
This current is comfortably determined with the vated lock-out function requires an acknowledge-
help of DIGSI during commissioning and entered ment of the TRIP command via the acknowledge-
into the protection relay parameters. Because no ment button on the relay or by the setting of a
differential voltage and thus no differential cur- binary input, e.g. by using an external switch.
rent occur in flowing currents during normal op-
eration, monitoring of the pilot-wire connection
is strongly advised. Audio frequency signals are
modulated onto the connection line.
Intertripping, remote tripping

In the event of protective tripping at the local end,
LSP2746.tif
an intertrip signal can be sent to the remote end

(using the same transmission equipment) in order
to isolate the faulted line. The remote tripping sys- Fig. 7 Settings for “Power system data”
tem, in which a signal coupled via binary input is
interpreted as a shutoff command for the circuit- 3.2 Line differential protection
breaker on the remote end, works according to the 3.2.1 Line differential protection
same pattern. Here too an audio frequency signal
As with all protection functions, the differential
is transmitted to the partner relay, as with inter-
protection can be switched either on or off at this
tripping.
point in order to simplify function-selective test-
ing. Additionally, there is the option to set these
n 3. Settings
parameters to “indication only”, so that for exam -
The parameter settings of both relays of the differ- ple at the time of commissioning all indications
ential protection system differ in only a few for this protection function are logged, but no
points. This is the reason why only the 7SD600 tripping occurs. The differential protection func-
settings of one line end are explained initially. tion must of course be switched on for normal
The differences are explicitly listed towards the operation. Regarding the differential protection
end of this chapter. function, only the tripping threshold IDIFF> must
be set (referred to the line rated current).

Referred to the summation current, this value 3.2.5 Intertrip function

must thus lie below the minimum short-circuit Normally, tripping is effected at both stations as a
current but above the inrush current and the result of current comparison. Tripping at one end
transformer faults of the primary and summation only can occur when an overcurrent release is
current transformers, taking the weighting factors used or with short-circuit currents only slightly
for the various fault types into account. The preset above the tripping value. Circuit-breaker inter-
value of 1.0 ⋅ I /IN,Line has proved over many years tripping can be parameterized in the unit with in-
to be a stable empirical figure. Taking account of a tegral pilot-wire monitor, so that definite tripping
weighting factor of more than 2 for single-pole at both ends of the line is assured.
faults, this corresponds to barely five times an ac-
cepted charging current of 10 % referred to IN,Line. 3.2.6 Differential protection blocking
Should 5 times the charging current be above this (spill current)
value, IDIFF> must be increased. This differential current monitoring function
This charging current is calculated according to reacts to a permanently low differential current,
the equation: which can be produced by phase failure at the
IC = 3.63 · 10-6 · UN · fN · CB’ · s summation current transformer (e.g. due to wire
break) and blocks the differential protection func-
IC = Charging current to be ascertained in A primary
UN = System rated voltage in kV
tion. The threshold (parameter 1550) is set slightly
fN = System rated frequency in Hz over the capacitive losses of the pilot wires, which
C B’ = Operating capacity of the line in nF/km at a power system frequency of 50 Hz can be
s = Line length in km estimated according to
Ispill (%) = 0.025 ⋅ IN, Line ⋅ lLine (km)
3.2.2 Blocking with second harmonic
A transformer may also be situated within the With a line length of 12 km, this gives a value of
7SD600 protection range. However, in this case 0.3 ⋅ IN, line for the spill current. In order to prevent
the transformer ratio must be recreated with ex- spurious tripping, the parameter is set at
ternal matching transformers, so that the current 0.4 ⋅ IN, line. In accordance with the default (preset)
magnitude and phase angle of the summation value, monitoring is delayed by 5 seconds.
current transformer inputs correspond on both
sides of the transformer. For this application it is
necessary to stabilize the differential protection in
relation to the transformer inrush. Because no
transformer is located within the protection range,
blocking of the differential protection is deacti-
vated by means of the second harmonic. Conse-
quently, the tripping threshold set values for both
the second harmonic and the maximum differen-
tial current (which is blocked by this function) are
LSP2658en.tif
irrelevant.
3.2.3 TRIP delay Fig. 8 Setting the differential protection function

In certain applications (e.g. reverse interlocking),
it can be necessary to delay the differential protec- 3.3 Pilot wire monitoring
tion somewhat. This delay can be set at this point. Pilot wire monitoring is extremely important for
monitoring the capability of the differential pro-
3.2.4 Local current threshold tection system. Since in fault-free operation, par-
A local current threshold (which must be ex- ticularly where operating currents are low, no
ceeded) can be set as a further tripping condition appreciable differential current occurs (due to
at the local end. With the preset value 0 ⋅ IN, Line transformer and measurement inaccuracies), wire
the protection relays trip at both ends if the differ- break or short-circuit would not be noticed,
ential protection responds. The local current which would lead to protection malfunction. Thus
threshold can be raised if, for example, in the case the pilot-wire function will be activated and the
of single-side infeed, the remote end (from which reaction of the protection will be defined. When a
no current is feeding onto the fault) should not be connection fault is recognized, the differential
tripped. protection can be blocked or the fault can simply
be reported – after an adjustable delay time.

In order to begin the communication between To avoid an overfunction, tripping can be delayed.
both relays in a defined manner, the station iden- Hence, a transient signal will not lead to a misin-
tification must be set differently. One 7SD600 is terpretation. In order to ensure reliable tripping
parametrized as master, the other as slave. even when a short signal is received, the transfer
trip signal can also be delayed until the circuit-
breaker opens. The preset values are intended to
ensure remote tripping.
3.6 Relay at the remote end
The parameter settings of the second 7SD600,
LSP2659en.tif
which is installed at the remote line end, are to the
greatest possible extent identical to the relay de-
Fig. 9 Settings of the pilot-wire monitoring for the local
scribed here. The identical summation current
relay (master) transformer connection and the same set value for
the line rated current are essential. All further pa-
3.4 Overcurrent-time protection rameters can normally be taken over by the local
Emergency overcurrent-time protection can be relay; however, under the key term “pilot-wire
activated either in the event of a recognized pilot- monitoring”, the station identification of “mas -
wire fault or generally when the differential pro- ter” has to be changed to "slave” so that this pa -
tection is deactivated. The local summation cur- rameter is set differently in both relays.
rent must likewise be used as a measuring quantity
and the setting must be referred to the line rated
current. Here too, when setting the current thres-
hold, the weighting factors for the various fault
LSP2660en.tif
types must also be taken into consideration. The

current thresholds are set – as far as possible – be-
tween maximum operating current and minimum Fig. 11 Settings of the pilot-wire monitoring for the relay at the
short-circuit current. The associated delay time is remote end (slave)
adjusted to the network grading plan as closely as
possible, in order to maximize selectivity. n 4. Summary
Undelayed and at the same time rigorously selec-
tive protection of cables and lines reduces the
consequences of unavoidable power system dis-
turbances. For one this means protection of the
LSP2692.tif
equipment, and secondly a contribution to a max-

imized level of supply security.
Fig. 10 Settings of the pilot-wire monitoring A differential protection system consisting of two
“Emergency overcurrent-time protection”
SIPROTEC 7SD600 relays and the associated
summation current transformers offers compre-
3.5 Remote tripping/Transfer trip
hensive protection of cables and overhead lines.
A signal injected via binary input (e.g. from the Extensive additional functions allow trouble-free
circuit-breaker failure protection) can be trans- connection of the relays and integration into com-
mitted as an audio-frequency signal to the remote plex power system protection grading.
end through the pilot wires in order to effect trip-
ping of the circuit-breaker there. Thus reliable The default settings of the relay are selected in
transmission is ensured even when pulses are very such a way, that the user only has to configure the
short. The transmission duration can be pro- known cable and primary transformer data. Many
longed. The protection relay at the remote end of the preset values can be taken over with no
must be appropriately set for the reception of a problem and thus substantially reduce effort in-
remote tripping signal, i.e. the “transfer trip” volved in parameterization and setting.
function must be switched on.

Differential Protection of
Cables via Fiber Optics
(Relay Type: 7SD610)
n 20 kV single-core cross-linked polyethylene ca-

ble N2XS(F)2Y 1x120RM/16
n Resonant-earthed system
n Fiber-optic link
n ANSI 87L differential protection
n ANSI 50/51 definite-time overcurrent-time
protection as backup protection
n ANSI 50HS instantaneous high-current
switch-onto-fault protection
n ANSI 49 thermal overload protection
n 1. Introduction
The ever higher load imposed on primary equip-
ment requires it to be protected in a selective
manner and fast fault clearing in case of a short-
circuit, in order to minimize possible consequen-
tial damage resulting from faults. For overhead
lines and cables this requirement is met by line
differential protection relays.
LSP2638.tif
A full example of how to set SIPROTEC 4 7SD610
protection relays for a power cable in the distribu- Fig. 1 SIPROTEC 7SD610 line differential protection relay
tion network is described, in addition to notes on
design.
The 7SD610 numerical differential protection re-
lay is a modern short-circuit protection relay for
cables and overhead lines in power supply system.
Due to rigorous local selectivity – the protected
zone is limited at both ends of the line section –
power system topology and voltage levels play no
role. Furthermore, the star-point conditioning of
the current network is of no significance as cur- Fig. 2 Differential protection for a line
rent comparison takes place per phase and thus
variable weightings for different faults – as they
occurred in the conventional summation current Each 7SD610 compares locally measured current
transformer differential protection process – are values with those from the remote end and de-
nowadays unimportant. Due to its selectivity, the cides independently whether there is a system dis-
differential protection is generally set as an unde- turbance or not. A communication link between
layed, instantaneous main protection since no both relays is required to exchange the measured
other protection can disconnect the line more values. The relays are designed for a fiber-optic
quickly and selectively. link, which is the preferred method. System-tested
communication converters for other transmission
2.1 Differential protection (ANSI 87L) media (copper conductors, ISDN line, digital
The differential protection function of the 7SD610 communication networks with X21 or G703.1)
detects short-circuits using phase-selective com- are nevertheless also available.
parison of the current values measured by sepa-
rate relays at both ends of the line in the zone to
be protected, including weak current or high-re-
sistance short-circuits.
The differential protection function implemented In order to ensure reliable operation of the differ-
in the 7SD610 uses two algorithms in order to ential protection system, the current transformers
meet the demands of speed and sensitivity. The deployed must comply with the following
charge comparison process integrates the meas- requirements:
ured currents and compares the charge values at
n 1st condition:
both ends over a short time interval. This simple
When the maximum short-circuit current is
process makes it possible to detect high-current
flowing through, current transformers may not
faults as quickly as possible.
be saturated in steady state.
This crude algorithm is complemented by a sub- I
stantially more sensitive vector comparison pro- n' ≥ Kd max
I N prim
cess. In this process, the current vectors per phase
of both relays are compared with each other at n 2nd condition:
each sampling time point. In particular, the errors The operating overcurrent factor n' must be at
for each measured value during the process are least 30 or a saturation-free time t'AL of min ¼
considered. This includes the current measured period is ensured
value error based on the stored transformer data. n' ≥ 30 or t'AL ≥ ¼ period
The error consideration also takes into account
both the signal transmission time of the mea- n 3rd condition:
sured-value telegram from the partner relay and Maximum mutual ratio of the current trans-
the cable charge current. Finally, each protection former primary rated currents at the ends of the
relay can decide whether its own (directly mea- object to be protected
sured) current value corresponds to the measured I prim max
≤8
value received by telegram from the remote end I prim min
via the functional interface, including all magni-
tude and phase errors. If this is not the case, a fur- 2.2 Backup protection functions
ther (fault) current must be responsible for the As usual with modern, numerical protection
difference and the 7SD610 decides on tripping. relays, the 7SD610 also offers a range of further
Converted to the conventional characteristic, this protection and additional functions, which make
means that the restraint current is not simply it flexibly customizable for almost all uses. The
formed as the sum of the magnitudes of the cur- user must nevertheless be aware of the lack of
rents measured at both ends, but as the sum of the hardware redundancy when deploying these func-
errors described above plus the minimum tripping tions. For this reason at least one additional, sepa-
threshold IDIFF>, which are converted into a single rate short-circuit protection relay should be
current component. Protective tripping occurs at installed. Depending on voltage level and/or im-
the moment when the differential current is grea- portance of the line, this can be a separate distance
ter than the adaptively formed restraint current. protection (7SA6) or a definite-time overcurrent-
time protection relay (7SJ6).
The overcurrent-time protection included in the
7SD610 should therefore only be used as backup
protection against external faults in the power
system outside the differential protection zone.
2.2.1 Overcurrent-time protection (ANSI 50/51)

Parameterization makes it possible to specify
whether the three-stage overcurrent-time protec-
tion included in the 7SD610 should be working
permanently as an independent protection func-
tion (backup), or whether it should only be acti-
vated as an emergency function in case of mal-
Fig. 3 Differential protection tripping characteristic function in the communication link.
As mentioned above, if the backup function is used,
the concept of hardware redundancy must not be
neglected. Consequently, the backup definite-time
overcurrent-time protection function is recom-
mended primarily for protection outside the differ-
ential protection area, such as e.g. protection of an
incoming feeder panel in a substation.

Two of the three stages (I> and I>> stage) are n The breaker failure protection in the 7SD610
configured as incoming feeder protection in this has two stages. If a TRIP command issued by a
case. It has only been possible to authorize and set protection relay does not lead to the fault cur-
the I>>> stage in such a way, that it trips high- rent being shut off, the 7SD610 can initially re-
current faults quickly in this exceptional situation, peat the TRIP command before, at the second
losing the selectivity. If the overcurrent-time pro- stage, the higher-level protection is informed of
tection is used as an emergency function, all stages this malfunction by parallel wiring and trips the
can be set in terms of tripping threshold and delay circuit-breaker allocated to it.
time for this exceptional case, in line with selectiv- n The 7SD610 supports three-pole and sin-
ity and speed. gle-pole circuit-breaker activation, required
particularly frequently on the high-voltage level,
2.2.2 Instantaneous high-current switch-onto- thanks to its phase-selective operation.
fault protection (ANSI 50HS) n Transformers and shunt reactors in the differ-
This function is meant to disconnect immediately ential protected zone are also governed by inte-
in the event of single-end switching onto a high- grated functions.
current short-circuit. n By means of a connected binary input, a TRIP
The measured values of each phase, filtered to the command can be generated by the 7SD610 via
fundamental component, are compared with the an external coupling.
set threshold. n The digital communication link of the R2R
If the measured value exceeds twice the threshold, interface makes it possible to transfer 4 remote
protective tripping occurs immediately. For this commands and 24 remote messages from one
function, the circuit-breaker position of the re- relay to the other and process them individually
mote end must be known. in that relay.
A further stage of this protection function works n Because the 7SD610 also has voltage inputs, the
without data on the status of the circuit-breaker at line-to-earth voltages of the three phases and,
the remote end, but can only be used if current where applicable, the shift voltage can be con-
grading above the object to be protected is possible. nected to the relay. This does not affect the
protection function, but makes it possible to de-
2.2.3 Thermal overload protection (ANSI 49) tect the measured voltages and to calculate with
The thermal overload protection prevents over- the current measured values the derived electri-
loading of the object. In the case of the 7SD610, cal magnitudes such as active power, reactive
this function is used specifically for a transformer power, apparent power, cos ϕ (power factor)
situated within the protected zone, but is also ap- and frequency.
propriate for power cables that are working to full
capacity. n 3. Setting example:
As an example, the settings of the 7SD610 relays
The 7SD610 uses a thermal model to calculate
are described, such as are intended to protect a
(from the measured phase currents and from the
20 kV single-core XLPE cable of type N2XS(F)2Y
set parameters that characterize the object to be
1x120RM/16 with a length of 9.5 km. The cable
protected) the temperature of the equipment. If
rated current is 317 A, on side 1 a new 400 A/1 A,
this temperature exceeds an adjustable threshold,
10P10, 5 VA current transformer is used, and an
the 7SD610 issues a warning message, and if a sec-
existing 300 A/5 A, 10P20, 30 VA current trans-
ond, higher threshold is exceeded, the protection
former is located at the remote end. The maxi-
trips.
mum short-circuit current flowing through is
12.7 kA.
2.3 Additional functions
The additional functions listed in the following
are not used in the example given and are there-
fore only mentioned for the sake of completeness.
n Automatic reclosure makes it possible to
quench arc short-circuits on overhead lines by a
brief interruption of the current flow, i.e. not
necessarily immediately and fully disconnecting
the line.

3.1 Checking the transformers n’ = 44.6

Initial checks must be made to ascertain whether and for transformer 2
or not the transformer requirements are met. The
quotient of the primary side transformer rated n’ = 97.1.
currents evidently amounts to 400 : 300; a value These values must now be greater than or equal
smaller than 8 is thus achieved. to the required overcurrent factors in order to be
The operational overcurrent factor is calculated able to transmit the maximum through-flowing
from the formula short-circuit current at a level of 12.7 kA in a
saturation-free manner.
PN + Pi
n' = n ⋅ The following are required for transformer 1
P'+ Pi
n’ > 12700 A/400 A = 31.75
Equation 1
n’ = Operational accuracy limiting factor and for transformer 2
n = Rated overcurrent factor
n’ > 12700 A/300 A = 42.33.
PN = Rated burden of the current transformers [VA]
Pi = Inherent burden of the current transformers [VA] In both cases this is clearly achieved, as is the third
P’ = Actual connected burden [VA] condition n’ > 30. Consequently, the transformers
The current transformer’s own inherent load is available are suitable for use in this differential
calculated from protection system.
Pi = Ri · IN,CT2 3.2 Local relay settings

Equation 2 The parameter settings of both the differential
protection relays usually differ only in a few
If Ri (the transformer’s secondary winding inner points. This is why only the settings mentioned at
resistance) is unknown, the estimation of the beginning of this application example of a
Pi = 20 % ⋅ PN is a good approximation. To arrive 7SD610 are explained initially. The settings to be
at the actually connected burdens, all burdens changed of the second relay are explicitly listed
connected to the transformer core must be added. towards the end of this chapter. Initially, using
In this example, it is assumed that only the bur- the DIGSI 4 software for configuration, the 7SD6
dens of the protection relay (0.05 VA for relay relay with Order No. 7SD6101-4BA39-0BA0+M2C
rated current of 1 A, 0.3 VA for relay rated cur- is applied and opened in the current project.
rent of 5 A) and the incoming feeder cable load
are concerned. 3.2.1 Functional scope
The latter is calculated from the formula In the next step the parameters, beginning with
the “functional scope”, are set. At this point, we
2 ⋅ ρCu ⋅ l Line
PLine = determine which of the functions provided by the
aLine protection relay should be used. The other func-
Equation 3 tions are set at “not available” and are conse -
quently concealed for the remainder of the relay
PLine = Incoming feeder cable load [VA] parameterization. For our example, we have se-
ρCu = Specific resistance of Cu [ 0.0175 Ω mm²/m]
lLine = Secondary-side, single line length [m]
lected differential protection as the main protec-
aLine = Line cross-section [mm²] tion function together with overcurrent-time
protection, instantaneous high-current switch-
It is clear from equation 2, that with a second- onto-fault protection and overload protection.
ary-side transformer rated current of 5 A, an in-
coming feeder cable load 25 times higher appears
than for 1 A.
For our example, a 5 m incoming feeder cable
(single distance, therefore factor 2) with a cross-
section of 4 mm² is assumed. From this we cal-
culate an incoming feeder cable load for both
transformers of 0.045 VA (transformer 1) and
LSP2661.tif
1.116 VA (transformer 2).

The overcurrent factors for both transformers are Fig. 4 “Functional scope” menu item settings
now calculated from these values. According to
equation 1 for transformer 1 we have

3.2.2 Power system data Next, three parameters must be set. These charac-
In the section called “power system data 1”, the terize the current transformer in terms of its char-
parameters defined by the primary equipment are acteristic progression and also define differential
set. These are in particular the current transfor- protection sensitivity.
mer transformation ratio (400 A/1 A), the posi- Broadly, due to transformer faults, the influence
tion of its star (neutral) point (assumed to be of current-proportional measuring accuracy can
“on the line side”) as well as the rated frequency be divided into two areas, which are separated on
(50 Hz) of the power supply system. the current scale by the quotients n’/n. These per-
centage error values are dependent on the trans-
former class and can be taken from the Table 1
below. The footnote demonstrates that the quo-
tient n’/n should be set at a maximum of 1.50,
which corresponds to a “defensive” setting that
LSP2662.tif
prematurely migrates to the higher error influence

and thereby increases the self-restraint.
Fig. 5 “Power system data/Transformer data” menu item
settings
The preset values were defined under the same as-
pect. They refer to a transformer in the “10P” class
and also lie “on the safe side” for the quotients.
On the next card the minimum and maximum
These preset values can be left for all current
circuit-breaker trigger times are input in order to
transformer types. Under certain circumstances,
ensure the execution of switching commands.
some of the potentially very high differential pro-
tection sensitivity may be lost.
For our example this means, that the presetting
for both error values are suitable; the quotient can
be increased to the set value of 1.50, since in this
case n’/n = 4.46 is clearly over the recommended
maximum value of 1.50.
Fig. 6 Current transformer fault approximation
Transformer Standard Error at rated current Error at rated Setting recommendations

class overcurrent
Transformation Angle factor Address 251 Address 253 Address 254
± 60 min ≤5% ≤ 1.50

1)
5P IEC 60044-1 1.0 % 3.0 % 10.0 %
≤ 10 % ≤ 1.50
1)
10P 3.0 % – 5.0 % 15.0 %
± 30 min ε ≤ 10 % ≤ 1.50
1)
TPX IEC 60044-1 0.5 % 1.0 % 15.0 %
± 30 min ε ≤ 10 % ≤ 1.50
1)
TPY 1.0 % 3.0 % 15.0 %
± 180 min ε ≤ 10 % ≤ 1.50
1)
TPZ 1.0 % 6.0 % 20.0 %
± 18 min (only I ∼)
≤ 1.50
1)
TPS IEC 60044-1 3.0 % 10.0 %
BS: Class X
≤ 1.50
1)
C100 to ANSI 5.0 % 15.0 %
C800
Table 1 Current transformer data setting recommendations
1) If n'/n ≤ 1.50, then setting = calculated value;

if n'/n > 1.50, then setting = 1.50.

1210 IDIFF>: Pickup value

The IDIFF> stage corresponds to the tripping
threshold of the actual current comparison pro-
tection and is set at approximately 2.5 times the
charge current. This charge current is calculated
LSP2663.tif
according to the equation:

IC = 3.63 · 10-6 · UN · fN · CB’ · s
Fig. 7 Settings in the menu item “Power system data
1-I- CT characteristic” IC Primary charge current to be calculated [A]
UN Power system rated current [kV]
When parameter set switchover is deactivated, fN Power system rated frequency [Hz]
C B’ Line operating capacity [nF/km]
only the “parameter set A” is available for the fur -
s Line length [km]
ther settings. Under “power system data 2”, only
the rated operating current of the line (317 A), as The data for the single-core oil-filled cable to be
well as correct line state recognition and connec- protected are: CB’ = 235 nF/km, s = 9.5 km
tion recognition details, are set. At this point the
At a rated voltage of 20 kV and a power system
rated operating current of the object to be pro-
frequency of 50 Hz, a charge current of 8.1 A is
tected (i.e. the power cable) in particular can devi-
calculated from the above equation. Conse-
ate from the transformer rated current. The rated
quently, a set value of 20.3 A (primary) arises for
operating current must be set identically for both
IDIFF> or, in the case of a current transformer ratio
7SD610 relays, since this value is the basis for the
of 400 A/1 A, a secondary value of 0.05 A. This
current comparison at both ends.
figure is below the minimum threshold setting of
0.10 A (secondary). The “safe side” was once again
3.2.3 Differential protection settings
sought with the preset value of 0.30 A. This figure
The differential protection is parameterized and results from the assumption of three times the
set in a few steps as the main 7SD610 protection charge current at a level of 10 %, referred to rated
function. As in the case of all protection functions current. In the case of transformers with compa-
included in the scope, the differential protection rable response qualities and which – in the event
can once again be switched either on or off at this of external faults – transmit the maximum
point in order to simplify function-selective through-flowing current while remaining unsatu-
checking. The differential protection function rated, this threshold can also be lowered to as little
must be switched on as a matter of course for as 0.10 A. With various transformer types (e.g.
normal operating status. iron-core and linear), the preset value remains
Concerning the differential protection function, unchanged, in order to ensure protection stability
only five parameters need be set in the example against transients in the event of external faults.
given. In particular, two discrete pickup thresh- In this case, on comparable transformers with
olds (IDIFF> and IDIFF>>) of the differential pro- good response (n’ high), a setting of 0.10 A or –
tection function are set. Both these values deter- with a safety margin – of 0.20 A is possible. If
mine the pickup thresholds of both protection al- 2.5 times the charge current is greater than 0.30 A,
gorithms (described above) of the differential pro- the higher value must clearly be set.
tection function.
When comparing the apparently very sensitive
1233 IDIFF>>: Pickup value preset value of 0.30 A with customary differential
The IDIFF>> value defines the charge comparison protection settings (IN), the different failure type
tripping threshold, which decides on tripping very weighting prevailing in the latter must be taken
quickly in the event of high-current faults. This into account. Single-pole faults are often detected
value is usually set at rated operating current. In by the summation current transformer circuit
the case of a resonant-earthed power system, the with a sensitivity higher by a factor close to 3,
setting must not be below the non-compensated which would also correspond to a IDIFF> threshold
earth-fault current. Otherwise the starting oscilla- close to 0.30 A.
tion on occurrence of an earth fault could lead to
(unwanted) tripping. Consequently, the Petersen
coil rated current provides a good guide for set-
ting the IDIFF>> threshold if this lies above the
rated line current.

3.2.4 Setting the communication

Both 7SD610 relays communicate via a fiber-
optic link laid parallel to the power cable. With a
section length of 9.5 km, a fiber-optic cable with
9/125 µm mono-mode fibers is used. This com-
LSP2664.tif
munication link also requires a few parameter
settings, which means that the presettings in the
Fig. 8 Settings in the menu item “Settings group A – “R2R interface” section can generally be left un-
Differential protection – Diff protection”
changed.
As already mentioned, data (i.e. predominantly
Further differential protection function the current measured values) is transmitted be-
parameters (settings) tween the two relays by telegram. Individual erro-
Three further parameters also exist for finer ad- neous or missing telegrams pose no problem since
justment of the differential protection function. they are counted for statistics purposes, but are
otherwise ignored. However, if such an error sta-
First, there is the possibility to raise the IDIFF> tus remains over long time periods and an initial
pickup threshold when the line is switched in. time threshold is exceeded, a link malfunction is
This is recommended when long off-load cables reported. At a second, higher threshold, this is rec-
or overhead lines are energized. In order to avoid ognized as a link failure (outage). It is also possi-
causing pickup of the differential protection in ble to set the length of time for which transmitted
this case, this parameter IDIFF> SWITCH ON remote signals should retain their “old” status
should be set at approximately 3.5 times the when a link malfunction is recognized.
charge current, provided that this value is greater
than IDIFF>. Tripping of the current comparison In the "CT 1” card file, the R2R interface is acti-
protection should only be delayed in exceptional vated and the type of communication connection,
cases and it is therefore advisable to leave the pre- in this case “fiber-optic cable direct”, is selected.
setting for T-IDIFF> unchanged at 0.00 s. In a reso- Further parameters can be left at the presettings.
nant-earthed system however, in the event of
single-pole pickup, a delay is recommended in or-
der to avoid tripping due to the earth-fault igni-
tion process. A delay of 0.04 seconds has proved
suitable.
Because it is not necessary to take a transformer
into account in the range of the differential pro-
tection system, the transient rush restraint may re-
main deactivated. All further parameter settings of
this small card are consequently irrelevant.

Under the concept of "differential protection to- 3.2.5.2 Overcurrent-time protection

pology” the relays must now be assigned the iden - Since further short-circuit protection is generally
tification number “n”. This differential protection provided in addition to the differential protection
system consists of two 7SD610 relays. One of the – for reasons of hardware redundancy in an inde -
two relays must be set as “relay 1”, the other as pendent relay – the integrated definite-time
“relay 2”. The difference lies in that the absolute overcurrent-time protection is only activated if
chronology management of the system conforms there is a communication link failure (outage).
to relay 1. Relay 2 adjusts itself accordingly and The current thresholds are set – as far as possible –
consequently the time data of both relays is always between maximum operating current and mini-
comparable. Since both the relays could also be mum short-circuit current. The associated delay
linked to each other via a digital communication time is adjusted to the power system grading plan
network in which more than one differential pro- in the best possible way, in order to maximize se-
tection system is communicating, each relay can lectivity. In the example here – without knowing
in addition have a relay identification number as- the minimum short-circuit current – the recom -
signed to it. This may only be used once in the mended setting is 20 % above the maximum per-
communication network. Both these addresses mitted continuous current for the cable (407 A),
must be set identically in both relays. In our ex- i.e. 488 A or 1.22 A secondary.
ample of a direct fiber-optic link, no adaptation of
the identification numbers is necessary. If current-dependent grading via the object to be
protected is possible, a high-current stage with
immediate disconnection can also be set. In this
case it must be ensured that the threshold does not
pickup in the event of fault current flowing
through.
If the definite-time overcurrent-time protection
integrated into the 7SD610 is set as a permanently
LSP2666.tif
active backup protection function, overcurrent

and high current stages can be used for regular
Fig. 9 Settings in the menu item “Settings group A – definite-time overcurrent-time protection duties
differential topology”
outside the differential protection area. In the
event of a communication link outage, the I>>>
3.2.5 Backup protection functions
stage can be used in the sense described above as
3.2.5.1 Instantaneous high-current an emergency definite-time overcurrent stage.
switch-onto-fault protection
This function is only operative if the cir- 3.2.5.3 Thermal overload protection
cuit-breaker at the remote end is open and the lo- Thermal overload protection prevents overload of
cal 7SD610 is informed of this via the communi- the object to be protected, in this case the 20 kV
cation link. Assuming that the function is acti- cable. Because the cause of the overload normally
vated in both relays, the data relating to the cir- lies outside the object to be protected, the over-
cuit-breaker position must also be detected by load current is a through-flowing current. The re-
the local 7SD610. For this purpose the data items lay calculates the temperature rise in accordance
00379 “>CB 3p Closed” and 00380 “>CB 3p Open” with a thermal single-body model according to the
in the allocation matrix in the “power system data 2” differential equation:
data group must be combined with the associated
binary inputs. The pickup threshold for the I>>> 2
dΘ 1 1  I 
stage should be set at the approximate charge cur- + ⋅Θ = ⋅ 
rent of the line. This value offers a sufficient safety dt τ th τ th  k ⋅ I N 
margin since the protection algorithm compares
the instantaneous values with double the set root- For each phase, the protection function calculates
mean-square (r.m.s.)value. The stage I>>>>, a thermal replica of the object to be protected
which is independent of the circuit- breaker, is left from the square of the phase current. The unfil-
deactivated (set value “ ∞”) as no current grading tered measured value is used so that harmonics
is possible via the object to be protected. are also taken into account in the thermal consid-
eration. Whether the overload function should in
fact disconnect when the tripping limit is reached,
or whether reaching this threshold should only be
reported, must initially be set.

The basic current for overload detection is the 3.3 Settings of the relay at the remote end
thermally continuous permitted current of the ob- The settings of the local relay just parameterized
ject to be protected (compare cable data). This can can mostly be used as a basis for parameter assign-
be referred to the protection relay rated current ment on the relay at the remote end. The record is
via the setting factor k: simply duplicated by copying and pasting. This
I max creates a new relay file, the only difference in
k= which is the VD address. This ensures that the re-
IN
cord copied belongs to another relay, even though
Consequently, at a maximum permitted continu- that relay was until now identical.
ous current of 407 A and a primary transformer In the “Settings group A – differential protection
rated current of 400 A, a value for k of 1.02 results. topology” section, the parameter 1710 must be set
The temperature rise time constant τth must also to “relay 2”. In the absence of this setting, there
be taken from the manufacturer’s data. It must be cannot be any communication between the two
borne in mind that this must be set in minutes, relays.
whereas often a maximum permitted 1 second
current is specified, and the same applies to our
cable. In this case the 1 second current is 17.2 kA.
The conversion formula is
LSP2668.tif
2
τ th 1  permissible 1 second current 
= ⋅ 
min 60  permissible continuous current  Fig. 11 Settings for the relay at the remote end in the
“Settings group A – differential topology” section
τth is 29.8 minutes in this case.
Before reaching the tripping threshold, a thermal The transformer data and the transformer charac-
and/or current alarm stage can be set. These teristic in the “Power system data 1” section must
should typically be set somewhere below the trip- be adapted. Whether intertripping should be op-
ping threshold in order to give the operating staff erable from both ends should also be checked.
sufficient time to reduce the equipment load. For Otherwise, this must also be changed.
the thermal alarm stage it is recommended that The settings in the “Power system data 2” (differ-
the preset value of 90 % be left unchanged. The ential protection function, R2R interfaces, instan-
current alarm stage is set somewhere below the taneous switch-onto-fault and overload) are iden-
maximum continuous permitted operating cur- tical for both relays and need not be changed. The
rent. 95 % of this figure is selected here, i.e. 387 A definite-time overcurrent-time function settings
primary. Referred to the transformer rated cur- are dependent on power system topology and
rent, this gives approximately 0.97 A secondary. must therefore be checked. If the relays are con-
Finally, it is also possible to set the method used to nected to a substation control system or RTU, the
calculate the temperature rise. This is calculated respective relay addresses must be checked.
separately for each phase. There is a choice as to
whether the maximum of the three excess temper-
atures (preset), the arithmetic mean of these three,
or the temperature rise calculated from the maxi-
mum phase current should be significant for com-
parison with the tripping thresholds. In this case,
the preset value remains unchanged, provided no
other algorithm must be preferred.
LSP2667.tif
Fig. 10 Settings in the menu item “Settings group A –

thermal overload protection”

n 4. Connection example n 5. Summary

Generally, connection of three phase current The instantaneous and at the same time selective
transformers to the 7SD610 in Holmgreen circuit protection of cables and lines reduces the conse-
is recommended in accordance with Fig. 12. This quences of unavoidable power system disturban-
allows the differential protection to work with the ces. For one this means protection of equipment,
three directly measured phase currents. For other secondly it contributes to maximizing supply
protection functions (e.g. definite-time security.
overcurrent-time protection), an earth current
A differential protection system consisting of two
summated from the three phase currents is avail-
SIPROTEC 7SD610 relays provides comprehen-
able. If there are higher demands for the accuracy
sive safeguarding of cables and overhead lines.
of the earth current, a core-balance current trans-
Built-in emergency and backup protection func-
former can also be connected to the 7SD610 IE in-
tions – as well as extensive additional facilities –
put (Fig. 13). In this case the modified transfor-
allow problem-free connection of the relay and
mation ratio for this input must be entered via pa-
integration in complex power system protection
rameter 221 in “Power system data 1 –
grading schemes, without the need for any
transformer data”.
additional equipment.
The preset values on the relay are selected in such
a way that the user only has to set the known cable
and primary transformer data. Many of the preset
values can be taken over without difficulty, there-
by reducing the effort involved in parameteriza-
tion and setting.
Fig. 12 Current transformer connection to three primary current transformers and

neutral point current (normal connection)
Important: The cable shielding must be earthed on the cable side.

Note: Changing over the current polarity (address 0201) also causes
a polarity reversal of the current input I4.
Fig. 13 Current transformer connection to three primary current transformers
and separate earthing transformer (core-balance current transformer)

Thermal Overload
Protection of Cables
n 1. Introduction
Failure of underground cables can be costly and
time consuming to repair. Protection systems are
designed to protect cables from the high current
levels present under fault conditions. However,
the temperature rise due to extended overload
conditions is just as likely to cause cable failure. As
the trend in power system operations is to utilize
equipment as close to operating limits as possible,
the importance of protecting equipment against
thermal overloads becomes more critical.
LSP2299-eps
Thermal overload protection calculates the tem-
perature of the conductor based on specific con-
BUIS085.eps
ductor data and the current present in the circuit,

and is used to protect conductors from damage
due to extended overloads. In this application ex-
ample thermal overload protection of under-
ground cables only is described.
Fig. 1
Thermal overload protection is normally used in
an alarm mode to notify system operators of the ∆Θ = temperature rise above ambient temperature
potential for cable damage. However, thermal ∆Θmax = maximum rated temperature rise correspond-
overload protection can be used to trip a circuit- ing to maximum current
breaker as well. In either case, the presence of τ = thermal time constant for heating of the
conductor
thermal overload can be detected and removed I = measured r.m.s. current based on the maximum
before cable failure occurs. rated overload current of the protected conduc-
tors: Imeas/Imax.
n 2. Thermal overload protection The solution to the thermal differential equation
Thermal overload protection with total memory is:
calculates a real time estimate of the temperature
−t
rise of the cable, Θ, expressed in terms of the max-  
Θ op = Θ amb + ∆Θ max 1 − e τ 
imum temperature rise, ∆Θmax. This calculation is  
based on the magnitude of currents flowing to the
load, and the maximum continuous current rating The initial value is Θamb, the ambient temperature
of the conductor. The calculation uses the solu- of the cable, and the steady state value is
tion to the first order thermal differential equa- Θamb +∆Θ, where ∆Θ is determined by the magni-
tion:
tude of I. The initial value, Θamb, is assumed to be
dΘ that temperature on which the cable ratings are
τ +Θ = I2
dt based. The steady state value is achieved when the
∆Θ temperature has reached its final value due to the
with = Θ = heating effects of I. At this point, the value of
∆Θ max
τdΘ/dt in Equation 1 is zero. Therefore, at steady
state, ∆Θ = ∆Θmax I2, where I = Imeas/Imax. The
transition between the initial value and the steady
state value is governed by the exponential expres-
−t
sion,1 − e τ . τ is a constant of the cable to be pro-

tected.

Fig. 2 shows the operating temperature of the ca- estimate a maximum continuous current based on
ble as a function of time and overload. With no conductor ampacity information. The ampacity of
load, the conductor is at its ambient temperature. conductors is specified based on circuit configura-
If an overload equivalent to the maximum rated tions, conductor temperature, and ambient tem-
current is added at some time, the temperature of perature. Also specified is the maximum operating
the cable will approach Θmax following the expo- temperature of the conductor, and correction fac-
nential tors for various conductor operating temperatures
−t and ambient earth temperatures.
1−e τ .
To determine the maximum continuous current
The conductor temperature due to a current over- rating of a cable, use the ampacity at the emer-
load, starting from no-load conditions, has the gency overload operating temperature, and not
same characteristic as shown in Fig. 2, with Θmax that of the maximum conductor operating tem-
becoming Θop, and IMAX becoming ILoad. However, perature. According to ICEA specifications, [1]
when the conductor already has some load pres- emergency overloads are permitted for only a total
ent, the characteristic of the operating tempera- of 100 hours per 12 month period, only for no
ture changes. The conductor will heat up the cable more than five such periods in the life of the cable.
to some steady state temperature. When an over- Therefore, it is desirable to trip or alarm for any
load is added, the final temperature of the cable is situation when the thermal overload reaches this
calculated as if the cable was at normal operating level. To determine the maximum continuous
temperature. However, the starting point of the current, remember that the conductor configura-
second (overload) characteristic will coincide with tion and ambient temperature effect the current
the steady state temperature of the normal load. rating.
This is illustrated in Fig. 3.
Example:
Circuit voltage: 12.47 kV
Cable size: 500 MCM shielded
copper cable
Conductor temperature: 90 °C
Ambient temperature: 20 °C
Configuration: 3 circuits duct bank
From conductor tables, the ampacity for 90 °C
copper conductor at 20° ambient temperature
with 3 circuits in duct bank is 360 amps. The
emergency overload operating temperature for
90 °C-cable is 130 °C. From Table 1, at 20 °C am -
bient temperature, the ampacity rating factor is
1.18.
Therefore,
Fig. 2 Temperature vs. time for an overload of Imax 360 A x 1.18 = 424.8 A maximum continuous cur-
rent
n 3. Calculation of settings
There are two required settings for thermal over-
load protection, the k factor, and the thermal time
constant τ. τ is specific to the properties of the ca-
ble. The k factor relates the maximum continuous
current rating of the cable to the relay.
3.1 Maximum continuous current of cable

The maximum continuous current rating of the
cable is used in determining the k factor setting,
and may be used in determining the setting for τ.
This current depends on the cross-section, insu-
lating material, cable design, and conductor con-
figuration. Cable manufacturers may specify the
maximum continuous current rating of their ca-
ble. If the rating is not available, it is possible to

3.2 Calculating the k factor

The k factor relates the operating current to the
relay to permit overload detection. The k factor is
defined as the ratio of the maximum continuous
current Imax to the rated relay current IN:
Example:
I max
k=
IN
Circuit voltage 12.47 kV

Cable size 500 MCM shielded
copper cable
Conductor temperature 90 °C
Ambient temperature 20 °C
Configuration 3 circuits in duct bank
Maximum continuous
Fig. 3 Temperature vs. time with an existing overload
current (Imax) 424.8 A primary
Current transformer Example for calculating the thermal constant τ:
ratio 800/5
Rated relay current (IN) 5 A secondary Circuit voltage 12.47 kV
Cable size 500 MCM shielded copper cable
424.8 Conductor
(800 / 5) temperature 90 °C
k= = 0.53
5 Ambient
temperature 20 °C
3.3 Thermal time constant Configuration 3 circuits in duct bank
The thermal time constant τ is a measure of the Maximum continuos
speed at which the cable heats up or cools down as current (Imax) 424.8 A primary
load increases or decreases, and is the time re- Maximum current
quired to reach 63 per cent of the final tempera- oür 1 second 35,975 A primary
ture rise with a constant power loss. The thermal 2
1 35,975
time constant is the determining factor for calcu- τ(min) = ⋅ = 119.5 min
lating the operating temperature as a percent of 60 424.8
the maximum permissible overload temperature,
as shown in Equation 2. τ may be available from Conductor Ampacity Max. continu- Short-time τ
the cable manufacturer. If no specification for τ is size ous current withstand (minutes)
available, it may be estimated from the permissible capability (1 sec)
short-circuit rated current of the cable, and the 1/0 160 189 7 585 27
maximum continuous current rating. It is com-
2/0 185 218 9 570 32
mon to use the 1 second rated current as the per-
missible short-circuit current. τ is calculated by 3/0 205 242 12 065 41
the following equation: 4/0 230 271 15 214 52
2 250 MCM 255 301 17 975 59
1 I 1 second
τ(min) = ⋅ 350 MCM 305 360 25 165 81
60 I max
500 MCM 360 425 35 950 119
If the short-circuit current at an interval other
750 MCM 430 507 53 925 188
than one second is used, the equation is multiplied
by this interval. For example, if the 0.5 second rat- 1000 MCM 485 572 71 900 263
ing is used: Table 1 Shielded copper conductor, 5001 - 35000 volts, 90 °C,
2
three-conductor cable, three circuits in duct bank
0.5 I 0 .5 second
τ(min) = ⋅
60 I max

2 2
I 
Θ op = Θ amb + ∆Θ max  meas  = 90 °+40 ° 
Conductor Ambient earth temperature 400 
temperature 
 I max   424.8 
°C 10 °C 15 °C 20 °C 25 °C 30 °C
= 125 ° C
75 0.99 0.95 0.91 0.87 0.82
Thus, for a load of 400 A, the conductor will heat
85 1.04 1.02 0.97 0.93 0.89
up to 125 °C.
90 1.07 1.04 1.00 0.96 0.93
100 1.12 1.09 1.05 1.02 0.98 n 4. Implementing thermal overload protection
105 1.14 1.11 1.08 1.05 1.01 Thermal overload protection in SIPROTEC relays
calculates the temperature for all three phases in-
110 1.16 1.13 1.10 1.07 1.04 dependently, and uses the highest of the three cal-
125 1.22 1.19 1.16 1.14 1.11 culated temperatures for tripping levels. Besides
130 1.24 1.21 1.18 1.16 1.13 the k factor and the thermal time constant, there
are two other settings for thermal overload pro-
140 1.27 1.24 1.22 1.19 1.17
tection. As seen in Figure 3, these are the “thermal
Table 2 alarm stage” and the “current overload alarm
Correction factors for various ambient earth temperatures
stage”.
Short-circuit rated current for copper
2
 I  ⋅ t = 0.0297 ⋅ log  T2 + 234 
n  
 A  T1 + 234 
I = Short-circuit current - Amperes [A]
A = Conductor area - circular mils
(0,001" diameter; convert to mm2 where
necessary)
t = time of short-circuit, seconds
T1 = Operating temperature 90 °C
T2 = Maximum short-circuit temperature 250 °C
Short-circuit current for aluminium

2
 I  ⋅ t = 0.0125 ⋅ log  T2 + 234 
n  
 A  T1 + 234 
LSP2588.tif
Table 1 lists calculated values of τ for common

conductors and configurations. Below Table 2 are
the formulas to calculate the short-circuit rating Fig. 4 Thermal overload protection settings
of conductors
4.1 Thermal alarm stage
3.4 Analysis of relay settings The thermal alarm stage sends an alarm signal be-
Combining the examples in Sections 3.1, 3.2, and fore the relay trips for a thermal overload condi-
3.3, the cable information leads to the following tion. The thermal alarm stage is also the dropout
relay settings: level for the thermal overload protection trip sig-
nal. Therefore, the calculated temperature must
k factor: 0.53
drop below this level for the protection trip to re-
τ: 119.5 minutes set. This stage is set in percent of the maximum
temperature. A setting of 90 % will meet most
The operating temperature at a given moment in
operating conditions.
time can be calculated using Equation 2.
Θamb = the conductor operating temperature = 90 °C.
∆Θmax = the maximum overload temperature
minus the initial operating temperature =
130 °C - 90 °C = 40 °C.
As shown in Section 2 above, to determine the
steady state temperature, the exponential term can
be replaced by I2, where I = Imeas/Imax. Based on
these settings, and a load current of 400 A,

Fig. 5 Logic diagram of the overload protection

4.2 Current overload alarm stage n 5. Summary

The current overload alarm stage sends an alarm The failure of underground cables due to heating
when the load current exceeds the value of the set- caused by long term overload conditions is easily
ting. This setting should be set equal to, or slightly prevented by using thermal overload protection.
less than, the maximum continuous current rating Based on information provided by cable manufac-
of the cable. turers, circuit configuration, and operating condi-
tions, it is simple to determine settings for thermal
4.3 Thermal overload protection as an alarm or overload protection. Thermal overload protection
tripping function is normally used to alarm for overload conditions,
Thermal overload protection may be configured to allow system operators to make informed deci-
to either “ON” (tripping) or “Alarm only” for sions on how to handle an overload to prevent ca-
overload conditions, and is normally set to ble damage. A thermal overload alarm, when
“Alarm only”. Configuring thermal overload pro - combined with a SCADA system, can be used to
tection to “ON” makes this a tripping function, track the amount of time the cable is exposed to
meaning it asserts the 0511 Relay TRIP function. overload, allowing for estimates of the remaining
The factory default configuration of the relay has life of the cable.
the 0511 Relay TRIP function set to close a contact
that trips the circuit-breaker. n 6. References
Insulated Cable Engineers Association Standard
“Alarm only” means thermal overload protection
P32-382-1994, “Short Circuit Characteristics of
does not assert the 0511 relay trip function. Ther-
Insulated Cable”, 1994, South Yarmouth, MA.
mal overload protection may still be programmed
to operate a binary output for tripping purposes. “Engineering Handbook”, The Okonite Com-
this configuration allows the thermal overload pany, 1999, Ramsey, NJ.
function to be used to warn operators of potential
H. Pender and W. Del Mar, “Electrical Engineer’s
cable failure due to overloads.
Handbook”, 4th Edition, Wiley & Sons, New
York, NY.
4.4 Adjusting settings for differences in ambient
temperature
The ambient temperature of the earth has a signif-
icant effect on the maximum continuous current
rating of the cable. In most applications, it is best
to assume one ambient earth temperature to per-
form relay setting calculations. However, in some
areas, there may be large seasonal differences in
ambient earth temperature. Using multiple set-
tings groups allows the relay to adapt the thermal
overload protection settings to large changes in
the seasonal temperature.
Changing the settings group can be accomplished
via binary input, remote command, or function
key, all of which require operator intervention to
accomplish the change. Another possibility is to
have the relay change settings groups based on
system conditions. Wiring the output of a temper-
ature sensor into an optional transducer input on
the 7SJ63 relay will permit the changing of settings
groups when the earth temperature passes a
threshold for a specified period of time.

Disconnecting Facility
with Flexible Protection
Function
n 1. Introduction
The flexible protection functions allow sin-
gle-stage or multi-stage directional protection to
be implemented. Each directional stage may be
operated on one phase or on three phases. The
stages may use optionally forward active power,
reverse active power, forward reactive power or
reverse reactive power as a measuring variable.
Pickup of the protection stages can take place
when the threshold value is exceeded or under-
shot. Possible applications for directional power
protection are listed in Table 1.
A practical application example for reverse-power
protection using the flexible protection function is
described below.
n 2. System example
2.1 Functions for the disconnecting facility
LSP2650.tif
Fig. 2 shows the example of an industrial
switchgear with autonomous supply from the il-
lustrated generator. All the lines and the busbar Fig. 1 Protection for industrial plants
shown are in three-phase layout (except the earth
connections and the connection for voltage mea- Direction Evaluation
surement on the generator). The two feeders 1 and
2 supply the customer loads. In the standard case, Overshoot Undershoot
the industrial customer receives power from the P Forward Monitoring of forward power Detection of motors running at
utility. The generator runs only in synchronous limits of equipment (trans- no load
formers, lines)
operation without feeding in power. If the utility
can no longer maintain the required supply qual- P Reverse − Protection of a local indus-
trial power system against
ity, the switchgear should be disconnected from feeding energy back into the
the utility power system and the generator should utility power system
assume autonomous supply. In the example − Detection of reverse energy
shown, the switchgear is disconnected from the supply from motors
utility power system when the frequency leaves the Table 1 Application overview, reverse-power protection
rated range (e.g. 1-2 % of the rated frequency), the
voltage drops below or exceeds a given value or
the generator feeds active power back into the
utility power system. Some of these criteria are
combined depending on the user’s philosophy.
This would be implemented using CFC.
Here, reverse-power protection with the flexible
protection functions is explained. Recommenda-
tions are given for frequency and voltage protec-
tion in the Chapter 4 on setting instructions.

Switchgear data
Rated power of generator SN, Gen = 38.1 MVA
Rated power of transformer SN, Transfo. = 40 MVA
Rated voltage of high-voltage side UN = 110 kV
Rated voltage of busbar side UN = 11 kV
Primary rated current of the current IN, prim = 2000 A
transformers on the busbar side
Secondary rated current of the current IN, sec = 1 A
Primary rated voltage of the voltage trans- UN, prim = 11 kV
formers on the busbar side
Secondary rated voltage of the voltage UN, sec = 100 kV
Table 2 Switchgear data for the application example
n 3. Protection functionality
The SIPROTEC protection relay 7SJ64 discon-
nects the switchgear from the utility power system
if the generator feeds energy back into the power
system (protection function P reverse >). This
functionality is achieved by using flexible protec-
tion. Disconnection also takes place in the event of
frequency or voltage fluctuations in the utility
power system (protection functions f<, f>, U<,
U>, Idir>., IEdir>/ 81, 27, 59, 67, 67N).
The protection receive the measured values via a
three-phase current and voltage transformer set
and a single-phase connection to the generator
voltage transformer (for synchronization). The
circuit-breaker CB2 is activated in the case of a
disconnection.
Fig. 2 Example of a switchgear with autonomous The transformer is protected by differential pro-
generator supply tection and inverse or definite-time overcurrent-
time protection functions for the phase currents.
2.2 System data In the event of a fault, the circuit-breaker CB1 on
A 110 kV line connects the switchgear to the util- the utility side is tripped by a remote link. Circuit-
ity power system on the high-voltage side. The cir- breaker CB2 is tripped additionally.
cuit-breaker CB1 belongs to the utility power
Overcurrent-time protection functions protect
system. The switch disconnector decouples the
feeders 1 and 2 against short-circuits and overload
switchgear from the utility power system if neces-
caused by the connected loads. Both the phase
sary. The transformer with a ratio of 10:1 trans-
currents and the zero currents of the feeders can
forms the voltage level to 11 kV. On the low-
be protected by inverse and definite-time over-
voltage side the transformer, the generator and the
current-time stages. The circuit-breakers CB4 and
two feeders are connected by a busbar. The cir-
CB5 are tripped in the event of a fault.
cuit-breakers CB2 to CB5 disconnect loads and
equipment from the busbar. The busbar could be equipped additionally with
the differential protection 7UT635. The current
transformers required for this are already shown
in Fig. 2.

3.1 Synchronization when connecting the

generator
In most cases the electricity customer is responsi-
ble for restoring the switchgear or substation to
normal operation after shutdown. The SIPROTEC
7SJ64 tests whether synchronous conditions are
satisfied. After successful synchronization, the
generator is connected to the busbar. The voltages
required for synchronization are measured at the
transformer and at the generator. The voltage at
the transformer is measured in all three-phase be-
cause this is also necessary for determining the di-
rection. The generator feeds the phase-to-phase
voltage U31 to the device input U4 via a voltage
transformer in a star-delta connection (see Fig. 3).
3.2 Connection diagram

Fig. 3 shows the connection of the relay for re-
verse-power protection and synchronization. The
power flow in positive or forward direction takes
place from the high-voltage side busbar (not
shown) via the transformer to the low-voltage side
busbar.
3.3 Reverse-power protection with flexible

protection functions
The reverse-power protection evaluates the active Fig. 3 Connection diagram for a 7SJ642 as reverse-power protection
power from the symmetrical fundamental compo- (flush-mounted housing)
nents of the voltages and currents. Evaluation of
the positive-sequence systems makes the reverse-
power determination independent of the asymme-
tries in the currents and voltages and reflects the
real load on the drive side. The calculated active
power value corresponds to the total active power.
In the connection shown in the example the po-
wer in the direction of the busbar from the relay is
measured as positive.
3.4 Function logic

The logic diagram in Fig. 4 shows the function
logic of the reverse-power protection.
The reverse-power protection picks up when the
configured pickup threshold is exceeded. If the
pickup condition persists during the equally para-
Fig. 4 Logic diagram of the reverse-power determination with flexible protection
meterizable pickup delay, the pickup message function
“P reverse pickup.” is emitted.
This starts the trip command delay. If pickup does After the blocking has been released, the reverse
not drop out while the trip command delay is run- power must exceed the pickup threshold and both
ning, the trip message “P reverse TRIP” and the times must run out before the protection trips.
time-out indication “P reverse timeout” are gen -
erated (latter not illustrated). The picked-up ele-
ment drops out when the value falls below the
dropout threshold. The blocking input “>P re -
verse block” blocks the entire function; i.e. pickup,
trip command and running times are reset.

n 4. Setting instructions 4.3 Undervoltage protection U< (ANSI 27)

4.1 Reverse-power protection The voltage dip in the event of a short-circuit in
The pickup value of the reverse-power protection the system is detected with the U< criterion. The
is set at 10 % of the generator rated output. In this U< criterion should always be linked with the
example the setting value is parameterized as sec- fault current direction, to open the coupler circuit-
ondary power in Watts. The primary and secon- breaker only in the event of a fault in the utility
dary power have the relation: system. The voltage level should be set to 0.5 x UN.
U N, sec I N, sec
Psec = Pprim ⋅ ⋅ 4.4 Parameterization of reverse-power
U N, prim I N, prim protection with DIGSI
A relay 7SJ64x (e.g. 7SJ642) is created and opened
With the indicated data, the pickup values can be
first in the DIGSI Manager. A flexible protection
calculated – taking account of Pprim = 3.81 MW
function 01 is configured for the given example in
(10 % of 38.1 MW) on the primary level – as
the scope of functions (Fig. 5).
100 V 1A
Psec = 381
. MW ⋅ ⋅ = 17.3 W
11000 V 2000 A
on the secondary level. The dropout ratio is

parameterized to 0.9. This gives a secondary drop-
out threshold of Psec, dropout = 15.6 W. If the pickup
threshold is reduced to a value close to the lower
setting limit of 0.5 W, the dropout ratio should
likewise be reduced to approximately 0.7.
The reverse-power protection requires no short
trip times as protection against undesirable feed-
back. In this example it is useful to delay pickup
LSP2652en.tif
and dropout by about 0.5 s and tripping by about
1 s. Delaying the pickup minimizes the number of
opened fault logs when the reverse power fluctu-
ates around the threshold value. If the reverse- Fig. 5 Configuration of a flexible protection function
power protection is used to make it possible to
disconnect the switchgear from the utility power Select “Additional functions” in the “Parameters”
supply system quickly in the event of faults in the menu to view the flexible functions (Fig. 6)
latter, it is advisable to select a higher pickup value
(e.g. 50 % of the rated power) and shorter delay
times.
4.2 Frequency protection f<, f>

The relay 7SJ64 contains 4 frequency stages. One
stage is parameterized as f> and set to 50.5 Hz; it
operates without time delay. This stage detects the
frequency increase caused by a short-circuit in the
utility. The other 3 frequency stages should be
parameterized as f< stages to serve as load shedd-
ing criteria for isolated operation of the industrial
power supply system.
Suggested settings:
LSP2653en.tif
f1< = 49.5 Hz t1 = 0.2 s

f2< = 49 Hz t2 = 0.1 s
f3< = 48 Hz t3 = 0.2 s Fig. 6 The flexible function is visible in the function
selection
On reaching stage f3<, the generator should be op-
erated in isolated mode to safeguard autonomous
auxiliary supply coverage.

First the function must be activated under “Set - 4.5 Configuration for reverse-power protection in
tings → General”, and “3-phase” operating mode DIGSI
must be selected (Fig. 7). The names of the messages can be edited in DIGSI
and adapted accordingly for this example. The
names of the parameters are fixed.
The DIGSI configuration matrix initially shows
the following indications (after selecting “Indica -
tions and commands only” and “No filter”:
Fig. 9).
LSP2654en.tif
LSP2656en.tif
Fig. 7 Selection of three-phase operation Fig. 9 Indications prior to editing
Select “Active power reverse” or “Overshoot” in

It is possible to edit short text and long texts to
the menu items “Measured quantity” and “Mea -
suit the application by clicking the texts (Fig. 10).
surement method”. If the “Display additional set -
tings” box is activated in the “Settings” menu
item, the threshold value, pickup delay and TRIP
command delay can be configured (Fig. 8). Since
LSP2657de.tif
the power direction cannot be determined in the
event of a measuring voltage failure, protection
blocking is useful in this case.
Fig. 10 Indications after editing
Configuring of the indications is performed in the

same way as for the configuring of the indications
of other protection functions.
n 5. Summary
With the flexible protection functions, it is easy to
implement measuring criteria not available as
standard, such as power direction. This measuring
LSP2655en.tif
criterion is a fully fledged protection function and

can therefore be integrated as an equivalent crite-
rion in a disconnecting facility. The synchroniza-
Fig. 8 Setting options of the flexible function
tion function of the SIPROTEC 7SJ64 can be used
to advantage here, to synchronize the industrial
power supply system to the utility power supply
system.


Earth-Fault Protection in
Systems with Isolated
Star Point
n 1. General earth-fault information

In a system with isolated star point an earth fault
is not a short-circuit, but an abnormal operating
state. It must be signalled and corrected as quickly
as possible. The way in which the earth fault is
identified depends on the configuration of the sys-
tem. In a radial system, sensitive earth-fault direc-
tion detection with sine ϕ measurement is the
LSP2367.eps
method; in a meshed system the transient earth-
fault relay is preferred. In the case of an earth fault
with no resistance, e.g. in phase L3, the voltage
UL3-E drops to zero and the voltages UL2-E and
UL1-E increase to the 3-fold value. A displace-
ment voltage UE-N appears. This is also referred to
a zero-sequence voltage (U0). In normal operation
it has the value of the phase-to-earth voltage. A Fig. 1 Transient earth-fault relay 7SN60
purely capacitive earth-fault current flows at the
fault location. This can create very unstable arcs.
In general, isolated systems operate up to a capaci- l Earth fault = no short-circuit
tive earth-fault current of 50 A. The UE-N is evalu- l Operation continues during single
ated for signalling the earth fault. earth fault
The U0 voltage can be calculated from the phase l Earth fault must be signaled and
voltages or it can be detected via the voltage trans- corrected as quickly as possible
former open delta winding (e–n delta). This wind -
ing generally has a greater ratio in the region of l Earth-fault location with watt-
factor 3. In the case of an earth fault, the measu- metric earth-fault direction
ring-circuit voltage is thus approximately 100 V. A measurement or transient earth-
voltage relay for earth-fault detection is set at 25 V fault relay
– 30 V, and a time delay of 5 s is appropriate. This Fig. 2 Isolated system
functionality is included in line protection relays
7SJ5.., 7SJ6.., 7SA5.. and 7SA6 depending on the
configuration chosen. If the relays are equipped
with three transformer inputs a phase-selective
earth-fault alarm can also be provided. U ≤ 40 V
serves as the criterion for recognizing the defective
phase and U ≥ 75 V for the fault-free phases.
Fig. 3 Voltages in normal operation

Earthing connection
must be led back
Fig. 4 Voltages for earth fault in phase L3 Fig. 7 Core-balance current transformer
a not excessively high ratio of the transformers

(<150/1 or 150/5). The second method, measure-
ment with a core-balance current transformer, can
be used for smaller earth currents. It delivers
better values for sensitive earth-fault detection. It
is important to ensure that the transformer is as-
sembled with precision. In the case of cut-strip
wound transformers it is essential that the core
surfaces lie directly on top of each other. It is also
critically important that the earthing connection
of the cable screen earthing is led back through the
transformer so that the sum of the phase currents
can actually be measured (see Fig. 6 and 7).
Fig. 5 Voltage transformer with open delta winding n 2. Sensitive earth-fault direction detection
with sine measurement
Radial system
Energy direction
3 ⋅ I 0 = IL 1 + IL 2 + IL 3
Core balance current

transformer
Fig. 6 Connection of currents Fig. 8 Radial system
Two methods can be used to measure the earth Earth-fault direction measurement is only appli-
current. cable in the radial system. If it is used in a meshed
system, meaningful results can only be expected
The Holmgreen circuit adds the three phase cur-
after switching over to radial lines.
rents (by means of appropriate connection of the
current transformers) and thus provides the earth
current. However, because each transformer has
always an error of measurement, this method for
measuring the earth current is only suitable in sys-
tems with higher earth-fault currents (> 40 A) and

Capacitive currents
Overhead line 20 kV ∼ 0.05 A/km
110 kV ∼ 0.30 A/km
Cable 10 kV ∼ 1.5 A/km
20 kV ∼ 3.0 A/km
110 kV ∼ 20.0 A/km
Earth-fault direction detection
The system capacitive current can be estimated by Forward
using the table or values given in cable manuals. Backward
Example to determine IE Fig. 9 Earth fault in radial system
Current
30 km 10 kV cable, 1.5 A/km Fig. 9 shows an example of how the indication of
IE = 1.5 A / km · 30 km = 45 A earth-fault direction detection could look in a spe-
cific case. It is important to note that not all the un-
Holmgreen-circuit, ratio of the affected circuits (or in the worst case scenario none
main current transformer 200/1 of them) indicate backward. If the partial current
Earth current at the protection relay 225 mA being delivered to the earth-fault location is lower
Setting IE> 150 mA than the limit value set, no direction indication oc-
In the case of an earth fault only the healthy parts curs. However, because of the voltage ratios, the
of the system continue to provide an earth-fault earth fault is recognized by all relays and the general
current; therefore the pickup value must always be earth-fault signal is given. For remote reporting the
lower than the maximum earth-fault current. In message “earth fault” must be transmitted once
an exact calculation the value of the longest line from the galvanically connected system. From the
section plus a safety margin must be subtracted individual feeders it is advisable only to transmit
from the maximum. the message “earth fault forward”. If the feeder with
“earth-fault forward” message is disconnected, the
Voltage settings: earth-fault message will be cleared.
Attention must be paid to the voltage settings as If the line affected by the earth fault is an open
follows: ring with several sectioning points, it is possible to
n Displacement voltage: value in the case of earth identify the earth-faulted section by moving the
fault: 100 V/ 3 isolating point.
n Measured voltage at the open delta winding
(e-n winding): value in the case of earth fault:

100 V Normal connection 1. Moving the isolating point 2. Moving the isolating point
n Threefold zero-sequence voltage 3U0: value in
the case of earth fault: 100 V· 3.

For connection to the e-n winding, a pickup value
of Ue-n> = 25 V is usual. For setting tripping val-
ues for the (calculated) displacement voltage,
25 V/ 3 is recommended. The proposed operat-
ing value for 3U0 is 25 V· 3.
Earth-fault report time delay: t = 5 s
Voltage setting for phase-selective earth-fault de-
Fig. 10 Searching for the earth fault in a ring system
tection:
Affected phase U ≤ 40 V
Example in Fig. 10:
Healthy phases U ≥ 75 V
Normally the isolating point is situated at C.
Type of measurement An earth fault has occurred and relay 8 has re-
Sine phi ported “forward”. If the isolating point is now
moved from C to A, which can be done (by load
Earth-fault detection
disconnection) with no interruption to supply, the
Signal only (disconnection following an earth
relay 7 indicates “forward”. The section A – C is
fault is not usual)
thus affected by the earth fault. If the isolating
point is now moved to B, the relay 8 once again
indicates “forward”. Thus section A – B is clearly
affected by the earth fault.
n 3. Earth-fault direction detection with

the transient earth-fault relay 7SN60
If the system is meshed, no clear direction indica-
tion can be obtained from the sine ϕ measurement.
The current direction in the case of an earth fault
cannot be definitely detected. Good locating results
are achieved using transient earth-fault relays.
These relay work with the charge-reversal process,
which occurs with the earth fault. The capacity of
the phase affected by the earth fault is discharged to
earth and the healthy phases are charged up to the
higher voltage value.
Forward This charge-reversal produces a large current,
Backward amounting to a multiplication (threefold or
fourfold) of the capacitive current. The transient
Fig. 11 Transient earth-fault relay indications
earth-fault relays are thus always connected to the
Holmgreen circuit.
It is important to be aware that the charge-reversal
process only occurs when the earth fault appears,
i.e. just once. Repeat measurements following
switching therefore have no meaning and lead to
confusion.
In order to identify the circuit affected by the
earth fault in a meshed system, an indication is
required from both ends of the line. Both relays
must indicate in a “forward direction”. It is there-
fore advisable to transfer the signals from the
transient earth-fault relay onto an image of the
system. It is then possible to decide quickly where
the earth fault is located.
In Fig. 11, the fault is located in the middle line
from ST 4 to ST 3, since here both relays are indi-
cating “forward”.
n 4. Summary
Operation can be continued when an earth fault
occurs in a power system with an isolated star
point. The fault can be located as described above.
The operator should quickly separate the fault
location from the system. Thus a double fault
(which – as a short-circuit – would cause a supply
Fig. 12 Transient earth-fault relay connection
interruption) can be avoided.
Fig. 13 Transients in an earth fault

Earth-Fault Protection in a
Resonant-Earthed System

In a resonant-earthed power system an earth fault
is not a short-circuit, but an abnormal operating
state. It must be signalled and corrected as quickly
as possible. The way in which the earth fault is
identified depends on the configuration of the
network. In a radial system, sensitive earth-fault
direction measurement with sine ϕ measurement
LSP2367.eps
is used; in a meshed system the transient earth-
fault measurement is preferred.
In the case of an earth fault with no resistance, e.g.
in phase L3, the voltage UL3-E drops to zero and
the voltages UL2-E and UL1-E increase to the
3-fold value. A displacement voltage UE-N accu-
mulates. This is also referred to as zero-sequence
Fig. 1 Transient earth-fault relay 7SN60
voltage (U0). Under normal operating conditions
it has the value of the phase-to-earth voltage.
The capacitive earth-fault current at the fault loca-
tion is compensated by the inductive current from l Earth fault = No short-circuit
the Petersen coil so that the active current at the l Operation continues
fault location is very small. A residual resistive
current remains and is determined by the ohmic l Earth fault must be signaled and
part of the coil. It is in the order of magnitude of corrected as quickly as possible
3 % of the capacitive coil current. The UE-N volt- l Earth-fault location with
age is evaluated for signalling the earth fault. watt-metric earth-fault direction
The U0 voltage can be calculated from the phase measurement or transient
voltages or it can be detected via the voltage trans- earth-fault relay
former open delta winding (e–n delta). This wind -
ing generally has a greater ratio in the region of
Fig. 2 Resonant-earthed system
factor 3. In the case of an earth fault, the mea-
suring-circuit voltage is thus approximately
100 V. A voltage relay for earth-fault detection is
set at 25 – 30 V, and a time delay of 5 s is appro -
priate. This functionality is included in line pro-
tection relays 7SJ5.., 7SJ6.., 7SA5.. and 7SA6
depending on the configuration chosen. If the re-
lays are equipped with three transformer inputs a
phase-selective earth-fault alarm can also be pro-
duced. U ≤ 40 V serves as the criterion for recog-
nizing the defective phase and U ≥ 75 V for the
fault-free phase.
Fig. 3 Voltages in normal operation

Earthing connection
must be led back
Fig. 4 Voltages for earth fault L3 Fig. 7 Core-balance current transformer
earthed system. The second method, the core-

balance current transformer is available for such
cases. This delivers definitely better values for
earth-fault detection. It is important to ensure
that it is assembled with precision. In the case of
cut-strip wound transformers it is essential that
the core surfaces lie directly on top of each other.
It is also critically important that the cable screen
earthing is led back through the transformer so
that the sum of the phase currents can actually be
measured.
Fig. 5 Voltage transformer with open delta winding n 2. Watt-metric earth-fault direction detection
with cos measurement.
Radial system
Holmgreen circuit
Energy direction
3⋅ I 0 = I L1 + I L2 + I L3
Core balance current

transformer
Fig. 6 Connection of currents Fig. 8 Radial system
Two methods can be used to measure the flowing Watt-metric earth-fault direction measurement is
earth current. only appropriate in the radial system. If it is used
in a meshed system, meaningful results can only
The Holmgreen-circuit adds the three phase cur-
be expected after switching over to radial lines.
rents (by means of appropriate connection of the
current transformers) and thus provides the earth
current. However, because each transformer has a
fault, this measurement method is not suitable for
the small residual resistive currents in a resonant-

Capacitive currents Fig. 9 shows an example of how watt-metric indi-

Overhead line 20 kV ∼ 0.05 A/km cation of the earth-fault direction detection could
look in a specific case. It is important to note that
110 kV ∼ 0.30 A/km
Cable 10 kV ∼ 1.5 A/km
20 kV ∼ 3.0 A/km
110 kV ∼ 20.0 A/km
The network’s capacitive current can be estimated

by using the table or values given in cable manuals.
Alternatively, the value can be read from the coil.
Example: Earth-fault direction detection
Forward
Current
Petersen coil with a rated current of 200 A,
momentarily adjusted to 180 A. Fig. 9 Earth fault in radial network
We can assume a capacitive earth-fault current of
180 A. If we estimate the residual resistive propor- not all the unaffected circuits (or in the worst case
tion at 3 %, 5.40 A is the result. This is trans- scenario none of them) indicate backward. If the
formed by the core-balance current transformer at active component of the partial current being de-
60:1, and 90 mA consequently arrive at the protec- livered to the earth-fault location is lower than the
tion relay. The pickup value should then be set at limit value set, no direction indication occurs.
approximately 50 mA. In the case of an earth fault However, because of the voltage ratios, the earth
only the healthy parts of the system continue to fault is recognized by all relays and the general
provide an earth-fault current; therefore the earth-fault signal is given. For remote reporting
pickup value must always be lower than the maxi- the message “earth fault” must be transmitted
mum earth-fault current. once from the galvanically connected system.
From the individual feeders it is advisable only to
Voltage settings:
transmit the message “earth fault forward”. If the
Attention must be paid to the voltage settings as feeder with “earth-fault forward” message is dis-
follows: connected, the earth-fault message will be cleared.
n Displacement voltage: value in the case of earth
fault: 100 V/ 3
n Measured voltage at the open delta winding Normal connection 1. Moving the isolating point 2. Moving the isolating point
(e-n winding): value in the case of earth fault:
100 V
n Threefold zero-sequence voltage 3U0: value in
the case of earth fault: 100 V· 3.
For connection to the e-n winding, a pickup value
of Ue-n> = 25 V is usual. For setting tripping
values for the (calculated) displacement voltage,
25 V/ 3 is recommended. The proposed pickup
value for 3U0 is 25 V · 3.
Earth-fault report time delay: t = 5 s Fig. 10 Searching for the earth fault in a ring system
Voltage setting for phase-selective earth-fault de-

tection: If the line affected by the earth fault is an open
Affected phase U ≤ 40 V ring with several sectioning points, it is possible to
Healthy phases U ≥ 75 V identify the earth faulted section by moving the
isolating point.
Type of measurement
Cos phi
Earth-fault detection
Signal only (disconnection following an
earth fault is not usual)

Example in Fig. 10:

Normally the isolating point is situated at C.
An earth fault has occurred and relay 8 has re-
ported “forward”. If the isolating point is now
moved from C to A, which can be done (by load
disconnection switches) with no interruption to
supply, the relay 7 indicates “forward”. The sec -
tion A – C is thus affected by the earth fault. If the
isolating point is now moved to B, the relay 8 once
again indicates “forward”. Thus section A – B is
clearly affected by the earth fault.
n 3. Earth-fault direction detection with
Forward the transient earth-fault relay 7SN60
Backward If the system is meshed, no clear direction indica-
tion can be obtained with watt-metric relays. The
current direction in the case of an earth fault can-
Fig. 11 Transient earth-fault relay indications
not be definitely detected. Good locating results
are achieved using transiente earth-fault relays.
These relays work with the charge-reversal
process, which occurs with the earth fault. The
capacity of the phase affected by the earth fault is
discharged to earth and the healthy phases are
charged up to the higher voltage value. This
charge-reversal produces a large current, amount-
ing to a multiplication (threefold or fourfold)
of the capacitive current. The transient earth-
fault relays are thus always connected to the
Holmgreen-circuit. It is important to be aware
that the charge-reversal process only occurs when
the earth fault appears, i.e. just once. Repeat mea-
surements following switching therefore have no
meaning and lead to confusion.
In order to identify the circuit affected by the
earth fault in a meshed system, an indication is
required from both ends of the line. Both relays
must indicate in a “forward direction”. It is there -
fore advisable to transfer the signals from the
transient earth-fault relay onto an image of the
system. It is then possible to decide quickly where
the earth fault is located.
In Fig. 11, the fault is located in the middle line
from ST 4 to ST 3, since here both relays are indi-
Fig. 12 Transient earth-fault relay connection cating “forward”.
n 4. Summary
Operation can be continued when an earth fault
occurs in a resonant-earthed power system. The
fault can be located as described above. The oper-
ator should quickly separate the fault location
from the system. Thus a double fault (which – as a
short-circuit – would cause a supply interruption)
can be avoided.
Fig. 13 Transients in an earth fault

Earth-Fault Protection in
a Low-Resistance-Earthed
System

In a low-resistance-earthed power system the
earth fault is a short-circuit and must therefore be
tripped by the short-circuit protection. In most
cases, in an impedance-earthed network (predom-
inantly medium voltage) the earth-fault current is
limited to a maximum of 2000 A. The minimum
short-circuit current occuring in the case of faults
at a greater distance is the standard pickup value
for earth-fault protection (IE>). It must be en-
sured that any earth fault is safely tripped. If
LSP2318..eps
short-circuit calculations for the power system are
carried out, it is useful to have the minimum
short-circuit currents calculated as well as the
maximum ones. These values less a safety margin
then form the basis of the IE> set value. The pick-
up value for the earth fault is lower than that for Fig. 1 SIPROTEC relay with earth-fault protection
phase failure, and in unfavorable cases can actual-
ly be below the rated current.
When distance protection relays (7SA5.., 7SA6.) l Earth fault = short-circuit
are used, the pickup value IE> merely acts to
l Disconnected by the short-circuit
release the phase-to-earth measuring systems. The
protection
same setting considerations nevertheless apply.
l Earth-fault detection is more sen-
n 2. Earth fault in an overhead line system sitive than phase current detection
Since an earth fault is by far the most frequent
fault in medium-voltage overhead line systems,
measures for its quick correction are welcome.
The most frequently used method is auto-
reclosure (AR). On the medium-voltage level
auto-reclosure is always triple-pole, and in the Fig. 2 Low-resistance-earthed system
case of high-voltage single-pole. Single-pole
auto-reclosing circuit-breakers are therefore a
precondition here. If an earth fault occurs on the
overhead line, the protection relays concerned will
pick up.
Following a TRIP command, there is a dead time
(500 ms for medium voltage, 1 s for single-pole
AR at high voltage), after this the line is recon-
nected. If the earth fault is corrected, operation
continues. If not, final disconnection takes place Fig. 3 Earth short-circuit in an overhead line system
within the set time. Approximately 70 % of earth
faults are corrected in this way without any major
interruption to operation.

Overcurrent-time protection relays 7SJ5.., 7SJ6..

can be ordered with this functional option, as can
the distance protection relays 7SA5.., 7SA6..
Settings for medium voltage:
Auto-reclosure Triple-pole
Dead time 500 ms
Action time 300 ms
Settings for high voltage (≥ 110 kV):
Auto-reclosure Single-pole
Dead time 1s
Action time 300 ms
n 3. Earth fault in a cable system

Auto-reclosure is not appropriate for a cable
system. In such cases, final disconnection is
unavoidable since an earth fault does not correct
itself. Selective grading avoids unnecessary
tripping in such cases.
n 4. Locating faults
It is possible to locate earth faults with the 7SJ6..
definite time overcurrent-time relay as well as
with the distance protection relays 7SA5.., 7SA6...
The precise setting of the line’s X layer is required.
This can be taken from the appropriate tables in
the cable manual. In contrast, the setting of the
ZE/ZL factor is more difficult. Only measurements
can reveal the true facts here. The relays measure
the impedance loop as far as the fault location.
If the above-mentioned set values are correct, an
accuracy of 3 % of the given fault distance can be
expected. In practice, the set values can be opti-
mized on the basis of an exact empirical analysis
of faults that have occurred.
n 5. Summary
In a low-resistance-earthed system an earth fault is
always a short-circuit.
The sensitivity of the pickup value for the earth-
fault protection has to be set adequately to reliably
trip upon each earth fault.
SIPROTEC line protection relays are available
with the earth-fault protection function as an
option.
In an overhead system about 70 % of the earth
faults are successfully eliminated by AR without
significant system interruptions.

Coping with Single-Phase

Load Diversity Using
Adaptive Relay Settings
n 1. Introduction
Setting ground overcurrent elements on the typi-
cal distribution circuit is a straightforward task.
The ground element setting must be sensitive
enough to operate for the coordination point with
the lowest ground (earth) fault current. However,
for circuits with large single-phase load diversity, a
large zero-sequence current (or 3I0), will be seen
in the neutral due to this imbalance. The ground
element pickup setting must also be high enough
that the relay does not trip for this level of zero-se-
quence current. The amount of 3I0 present will
change for different system conditions, such as
summer peak versus winter peak.
Typical relaying practices for a circuit that has
large single-phase load diversity involve making
some tradeoffs on the reliability of the protection
system. Generally, sensitivity (and therefore de-
pendability) is lowered so the relay allows the
maximum expected zero-sequence current in the
neutral (maintaining security).
This application note introduces the idea of adap-
tive relay settings to provide better ground
overcurrent protection for circuits with large sin-
gle-phase load diversity. The normal settings of
the relay use a ground element with maximum
sensitivity for fault conditions. As system condi-
tions change and the load diversity increases, the
relay adapts by decreasing the sensitivity of the
ground pickup, preventing a false operation by al-
lowing more 3I0 to flow. Adaptive relay settings
provide a cost-effective method of improving the
protection of distribution feeders, by automati-
cally maintaining both dependability and security
as system conditions change.
n 2. Current practices
For circuits with large single-phase load diversity, 2. Lower the ground element sensitivity to allow
there are three philosophies used for setting the maximum amount of 3I0 expected. This
ground relay elements: maintains some ground protection for all sys-
tem conditions and doesn’t require constant
1. Disable the ground element: This allows the correction of the relay settings. However, the
maximum amount of 3I0 present during load sensitivity of the ground relay is not ideal, espe-
diversity. However, the method provides mini- cially when load diversity is small. Like disabling
mum protection. Only the phase elements will the ground element, this method also lowers the
operate for a ground fault, and, due to load al- dependability of the protection relay to main-
lowance requirements, they may not have the tain security, but not as radically.
sensitivity to see all phase-to-ground faults. This
method greatly lowers the dependability of the
protection system to maintain security.

3. Change the ground element sensitivity for spe- When the amount of 3I0 decreases, and remains
cific system conditions. This method is used below the predetermined threshold, the relay re-
when load diversity increases greatly for a long turns to the normal settings group. This concept is
period of time, such as seasonal load peaks. illustrated in Fig. 1.
Protection is maximized as much as feasible,
Definition of the elements shown in Fig. 1:
but relay settings must be changed several times
a year. This method only lowers the depend- n Settings Group A
ability of the protection system when absolutely The “normal” settings group. The ground ele -
required, and maintains security. ment pickup setting is set with maximum sensi-
tivity to operate for a fault at the circuit reach
All three of these methods are in common use.
point, while allowing the amount of 3I0 present
From a pure protection standpoint, changing the
during normal single-phase load diversity con-
ground pickup settings for system conditions is
ditions.
the best method. This method provides the best
system reliability, by only lowering dependability
n Settings Group B
when required. The drawback is the cost in- The alternate settings group. The ground ele-
volved. With electro-mechanical relays, field ment is set to allow the amount of 3I0 present
changes of settings are required. Numerical relays during the maximum single-phase load diver-
generally require remote control, auxiliary relays, sity condition.
and operator action to change settings groups. n 3I0 Current Threshold
The level of zero-sequence current that indi-
n 3. New application concept cates a change in single-phase load diversity of
Adaptive relay settings allow a protection relay to the circuit. The value of this threshold must be
automatically, and independently, change settings large enough to indicate that the load diversity
as the system operating conditions change. The is actually changing, but less than the pickup
Siemens SIPROTEC 4 overcurrent relays (7SJ61, setting of the ground element for Settings
7SJ62, and 7SJ63), using the PLC programming Group A.
capability in the Continuous Function Chart n Threshold Timer
(CFC), can adapt relay settings automatically. For This time delay is used to ensure a larger load
this application, a relay will adapt the ground ele- diversity exists, and that the diversity is not a
ment sensitivity as the amount of zero-sequence transient condition. The increase in sin-
current changes due to increasing single-phase gle-phase load diversity, as indicated by the
load diversity. amount of 3I0, must the load diversity settings
group. The timer must be set long enough to
ensure the load diversity has increased to new
level, and has not just temporarily increased due
to a the addition of a large, transient, single
phase current draw condition. The timer must
also be set so the relay can differentiate between
fault conditions and changes in single-phase
load diversity.
n Reset Timer
This is the time necessary to indicate a decrease
in the single-phase load diversity. The single-
phase load diversity, as indicated by the amount
of 3I0, must stay below the 3I0 Current Thresh-
old for this length of time to return the relay to
the original settings group. This timer must be
Fig. 1 Adapting ground element settings set long enough to ensure the single-phase load
diversity has actually returned to normal condi-
To adapt the ground element settings, the relay is tion.
programmed to change settings groups as the
measured zero-sequence current changes over
time. As the amount of 3I0 increases, and remains
above a pre-determined threshold value, the relay
changes settings groups from one with maximum
ground element sensitivity, to one that provides
less sensitivity for higher phase diversity.

n 4. Application examples Element Setting Calculations

The load information shown in Table 1 is for a Setting Group A 0.5 Asec Reach = 1.5/0.5 = 3.0
circuit with a summer peak. As shown, there is a Ground element pickup Allowance = 0.5/0.26 = 192 %
significant increase in the load and the amount of Ground element time dial 1.0
3I0 during the summer. It is desirable to use peak
Setting Group B 1.1 Asec Reach = 1.5/1.1 = 1.36
demand values for load information, or when pos- Ground element pickup Allowance = 1.1/0.87 = 126 %
sible, historical data for the amount of zero-se-
Ground element time dial 1.0
quence current. For this example, the minimum
fault current that the ground element must oper- 3I0 current threshold 0.45 Asec
ate for is 1.5 Asec. Threshold timer 30 min
Winter Summer Reset timer 15 min
Phase A Load 1.25 Asec 2.38 Asec Table 2 Example relay settings
Phase B Load 1.54 Asec 1.62 Asec
n 5. Implementation of adaptive relay settings
Phase C Load 1.35 Asec 1.42 Asec
in a SIPROTEC 4 relay
3I0 Current 0.26 Asec 0.87 Asec
Table 1 Typical circuit load
Table 2 lists possible relay settings for the circuit

load information from Table 1. The ground ele-
ment pickup setting in Settings Group A is set to
provide a desired reach sensitivity to the mini-
mum ground reach point, while allowing some
percentage of the normal amount of 3I0. For de-
pendability, to ensure the ground element trips
for all faults in the zone of protection, minimum
reach is set in the range of 2/1 to 3/1. For security,
the ground pickup should allow 120 % to 150 % Fig. 2 Adaptive ground element logic
of the normal amount of 3I0.
The ground element pickup setting for Settings Implementation of the adaptive relay setting logic,
Group B is set to maintain security by allowing using the SIPROTEC 4 7SJ61, 7SJ62, and 7SJ63 re-
some percentage of the expected amount of 3I0. lays, is in the Continuous Function Chart (CFC)
A typical setting range is 120 % to 150 %. This set- portion of the relay. This is a graphical program-
ting is balanced against the desire to maintain mable logic controller programming tool that per-
some dependability by having some reach sensi- mits the use of standard and advanced logic and
tivity to the minimum ground reach point. control capabilities. Using the CFC, the logic of
Fig. 2 is implemented in the relay.
The 3I0 Current Threshold is set at a level high
enough to indicate a change in the single-phase The zero-sequence current measured by the relay
load diversity, but still below the ground element is compared to the 3I0 Current Threshold in the
pickup setting for Settings Group A. The 3I0 Cur- “Measurement” plan of the CFC. The output of
rent Threshold level, and the Threshold Timer, this comparison is used in the “Slow PLC” plan to
must also account for the rate of change of the start either the Threshold Timer or the Reset
zero-sequence current. If the threshold is too Timer. If the output of the comparison is logical 1,
high, or the timer too long, the zero-sequence cur- the Threshold Timer is started. Otherwise, the Re-
rent may increase over the ground element pickup set Timer is started. When the Threshold Timer
setting before a settings change occurs. Once the expires, a SR flip-flop is used to latch the “>Set
ground element is in pickup, no settings group Group Bit0” command. Asserting the “>Set
change is possible, and the ground element will Group Bit0” command changes the settings group
eventually trip. from A to B. When the Reset Timer expires, the
SR flip-flop is reset, deasserting the “>Set Group
The Reset Timer is set at a time long enough to Bit0” command, and changing the settings group
ensure that the single-phase load diversity has from B to A.
actually decreased, to prevent the relay from
“bouncing” between settings groups.

n 6. Summary
Adaptive relay settings provide a method for
matching the reliability of the protection system
to the actual system conditions. When looking at
this example of single-phase load diversity, the
protection engineer has typically had to give up
some of the dependability of the protection sys-
tem to maintain the security of the protection sys-
tem. Adaptive relay settings give the relay the
ability to maximize the dependability of the pro-
tection system, while not effecting security at all.
The idea of adaptive relay settings is a very power-
ful concept that can be extended to many applica-
tions. One possible application is to automatically
change relay settings for temporary increases in
circuit demand, such as during switching opera-
tions.
n 7. References
William D. Stevenson, Jr. Elements of Power
System Analysis, 4th Ed., McGraw-Hill Book
Company, New York, NY; 1982
Turan Gonen, Electric Power Distribution System
Engineering, McGraw-Hill Book Company,
New York, NY; 1986
DIGSI CFC Instruction Manual, Siemens Power
Transmission and Distribution,
Inc., Raleigh, NC; 1999
SIPROTEC Time-Overcurrent, Overload, And
Motor Protective Relay with Bay Controller 7SJ61
V4.0/4.1 Instruction Manual, Siemens Power
Transmission and Distribution,
Inc., Raleigh, NC; 1999

400 kV Overhead
Transmission Line
Protection
n 1. Introduction
This application example will guide the reader
through all the steps required to set the distance
protection functions for a typical transmission
line. Standard supplements such as teleprotection,
power swing, switch onto fault, directional earth-
fault protection, etc. are also covered.
n 2. Key functions applied:

n Distance protection (ANSI 21):
Quadrilateral characteristic
n Teleprotection for ANSI 21:
POTT
n Earth fault O/C (ANSI 67N):
IEC curves, directional

n Teleprotection for ANSI 67N:
Directional comparison
n Power swing blocking
n Weak infeed:
Echo and trip
LSP2589.tif
n Overcurrent protection:
Emergency mode
n Auto-reclose:
Fig. 1 Universal protection for OHL
1 and 3-pole, 1 cycle
n Synchronism check:
Sync. and async. closing

n Fault locator:
Single-end measurement
n 3. Single line diagram and power system

data
The required time graded distance protection
zones are:
Fig. 2 Single line diagram of protected feeder
Zone number Function Reach Time delay

Zone 1 Fast underreach protection for Line 1 80 % Line 1 0s
Zone 2 Forward time delay backup, overreach 20 % less than Z1 reach on Line 3 1 time step
Zone 3 Reverse time delay backup 50 % Z-Line 1 2 time steps
Zone 4 Not applied – –
Zone 5 Non-directional 120 % Line 2 3 time steps
Table 1 Notes on setting the distance protection zones

Parameter Value
System data Nominal system voltage phase-phase 400 kV
Power system frequency 50 Hz
Maximum positive sequence source impedance 10 + j100
Maximum zero sequence source impedance 25 + j200
Minimum positive sequence source impedance 1 + j10
Minimum zero sequence source impedance 2.5 + j20
Maximum ratio: Remote infeed / local infeed (I2/I1) 3
Instrument transformers Voltage transformer ratio (LINE) 380 kV / 100 V
Voltage transformer ratio (BUS) 400 kV / 110 V
Current transformer ratio 1000 A / 1 A
Current transformer data 5P20 20 VA Pi = 3 VA
CT secondary connection cable 2.5 mm2 50 m
CT ratio / VT ratio for impedance conversion 0.2632
Line data Line 1 – length 80 km
Maximum load current 250 % of full load
Minimum operating voltage 85 % nominal voltage
Sign convention for power flow Export = negative
Full load apparent power (S) 600 MVA
Line 1 – positive seq. impedance per km Z1 0.025 + j0.21 Ω/km
Line 1 – zero seq. impedance per km Z0 0.13 + j0.81 Ω/km
Line 2 – total positive seq. impedance 3.5 + j39.5 Ω
Line 2 – total zero seq. impedance 6.8 + j148 Ω
Line 3 – total positive seq. impedance 1.5 + j17.5 Ω
Line 3 – total zero seq. impedance 7.5 + j86.5 Ω
Maximum fault resistance, Ph - E 250 Ω
Power data Average tower footing resistance 15 Ω
Earth wire 60 mm2 steel
Distance: Conductor to tower/ground (midspan) 3m
Distance: Conductor to conductor (phase-phase) 5m
Circuit-breaker Trip operating time 60 ms
Close operating time 70 ms
Table 2 Power system and line parameters
Based on the source and line impedance, the fol- |Ztot| = |(10 + 80 ⋅ 0.025) + j(100 + 80 ⋅ 0.21)|
lowing minimum fault current levels can be calcu-
|Ztot| = |12 + j116.8|
lated for faults on Line 1:
|Ztot| = 117.4
U source
I fault = with Usource = 400 kV
3 ⋅ Z tot The minimum three-phase fault current is there-
fore:
If fault resistance is neglected then for 3-phase
400 kV
faults: I 3ph min =
3 ⋅117.4
Ztot = sum of positive sequence source and line
impedance (as only current magnitudes are being I 3ph min = 1967 A
calculated, only the magnitude of the impedance
is relevant)

If fault resistance is neglected then for single- The available functions displayed depend on the
phase faults: ordering code of the device (MLFB). The selection
made here will affect the setting options during
Ztot = 1/3 (sum of positive, negative and zero se-
the later stages. Careful consideration is therefore
quence source and line impedance)
required to make sure that all
the required functions are se-
. ) + j(200 + 80 ⋅ 0.81)|
|2 ⋅[(10 + 80 ⋅ 0.025) + j(100 + 80 ⋅ 0.21)] + (25 + 80 ⋅ 013
| Z tot | = lected and that the functions
3 that are not required in this
|Ztot| = |19.8 + j166.1| particular application are dis-
abled. This will ensure that
|Ztot| = 167.3 only relevant setting
alternatives appear later on.
The minimum single-phase fault current without
fault resistance is therefore: 103 Setting Group Change Option:
Only enable this function, if more than one
400 kV
I1ph min = = 1380 A setting group is required. In this example
3 ⋅167.3 only one setting group is used; therefore this
function is disabled.
If fault resistance is included then for single-phase
faults: 110 Trip mode:
On OHL applications, single-pole tripping is
Ztot_R = Ztot + RF
possible if the circuit-breaker is capable of
|Ztot_R| = |RF + Ztot| this. The advantage is that during a single-
pole dead time the OHL can still transport
|Ztot| = |250 + 19.8 + j166.1|
some power and reduce the risk of system
|Ztot| = 316.8 instability. In this example both one and
three-pole tripping is used so the setting is
The minimum single-phase fault current with
1-/3-pole.
high resistance is therefore:
112 Phase Distance:
400 kV
I1ph min_ R = = 729 A As distance protection for phase faults is re-
3 ⋅ 316.8 quired, Quadrilateral must be selected. In
some cases (depending on the ordering code)
n 4. Selection of device configuration a MHO characteristic can also be selected.
(functional scope)
113 Earth Distance:
After selection and opening of the device in the
Here the earth fault distance protection char-
DIGSI Manager, the first step when applying the
acteristic is selected as for 112 above. There-
settings is entering the functional scope of the de-
fore set Quadrilateral.
vice. A sample screen shot showing the selection
for this example is given below: 120 Power Swing detection:
If power swing conditions can occur in the
vicinity of the applied relay, the power swing
detection must be enabled. It is required for
blocking of the distance protection during
power swings. At 380 kV it is common prac-
tice to Enable the power swing detection.
121 Teleprotection for Distance prot.:
To achieve fast tripping for all faults on the
circuit a teleprotection scheme must be ap-
plied.
LSP2590.tif
Fig. 3 Selected scope of functions

Parameter PUTT POTT Blocking Unblocking

Short line Not suitable as the Suitable as the Z1b set- Suitable as reverse Suitable as the Z1b set-
Zone 1 operation is es- ting may be substan- reach setting is inde- ting may be substan-
sential and Zone 1 set- tially larger than the pendent of line length tially larger than the
ting in X and R line impedance so that line impedance so that
direction must be signal transmission is signal transmission is
small on short lines secure for all faults on secure for all faults on
the line the line
Weak infeed Not suitable as the Suitable as the strong Partially suitable as the Suitable as the strong
Zone 1 operation is es- end detects all line reverse fault is also de- end detects all line
sential at both ends for faults with overreach- tected at the weak faults with overreach-
100 % line coverage ing Z1b. The weak infeed end but no trip ing Z1b. The weak
infeed end then echos at weak infeed end infeed end then echos
the received signal the received signal
Amplitude modulated Not suitable as the sig- Not suitable as the sig- Suitable as the signal is Not suitable as the sig-
power line carrier nal must be transmit- nal must be transmit- only sent when the line nal must be transmit-
ted through the fault ted through the fault is not faulted ted through the fault
location which attenu- location which attenu- location which attenu-
ates the signal ates the signal ates the signal
Frequency or phase Suitable as the signal Suitable as the signal Suitable as the signal Suitable as the signal
modulated power line can be transmitted can be transmitted can be transmitted un- can be transmitted
carrier through the fault loca- through the fault loca- der all conditions through the fault loca-
tion tion tion
Communication inde- Suitable Suitable Suitable Suitable
pendent of power line
Table 3 Selection of teleprotection scheme
In this case the selection is POTT

122 DTT Direct Transfer Trip: standard is Time Overcurrent Curve IEC in
If external inputs must be connected to initi- this application.
ate tripping via binary input, this function
131 Earth Fault overcurrent:
should be activated. The trip will then auto-
For high resistance earth faults it is advisable
matically be accompanied by the minimum
to not only depend on the distance protection
trip command duration (trip circuit seal in)
as this would demand very large reach settings
and event and fault recordings. In this exam-
in the R direction. The directional (and non-
ple this function is not required and there-
directional) earth-fault protection is very sen-
fore Disabled.
sitive to high resistance earth faults and is
124 Instantaneous High Speed SOTF therefore activated in this example. Here Time
Overcurrent: Overcurrent Curve IEC is selected.
When closing onto bolted faults extremely
132 Teleprotection for Earth fault Overcurr.:
large currents arise that must be switched off
To accelerate the tripping of the earth-fault
as fast as possible. A special overcurrent pro-
protection (activated under 131 above) a
tection stage is provided for this purpose. In
teleprotection scheme can be applied. In this
this example it will not be used and is there-
example a Directional Comparison Pickup
fore Disabled.
scheme will be applied.
125 Weak Infeed:
133 Auto-Reclose Function:
When weak infeed conditions exist (perma-
Most faults on overhead lines are of a tran-
nently or temporarily) at one or both ends,
sient nature so that the line can be energised
the weak infeed function must be Enabled.
successfully after fault clearance. For this
Refer also to Table 3.
purpose an automatic reclosure function can
126 Backup overcurrent: be implemented to minimise the line outage
When the distance protection is in service, it by reclosing with a set or flexible dead time.
provides adequate backup protection for re- In this application 1 AR-cycle will be applied.
mote failures. The overcurrent protection in
the distance relay is usually only applied
when the distance function is blocked due to,
for example, failure of the measured voltage
circuit (VT-fuse fail). This will be done in
this example, so the function must be acti-
vated. The selection of the response curve

134 Auto-Reclose control mode: The ‘advanced’ settings can in most cases be left
If, as in this example, single and three-pole on the default setting value.
tripping is used, the auto-reclose function is
triggered by the trip command. If the trip is
due to a backup protection operation (e.g.
Zone 2) then reclosure is normally not de-
sired. By application of an action time which
monitors the time between fault detection
and trip, reclosure can be prevented for time
delayed tripping (longer than set action
time). In this example the auto-recloser will
be triggered with Trip and Action time mon-
itoring.
135 Synchronism and Voltage Check:
LSP2591.tif
Before closing a circuit-breaker it is advisable
to check that the system conditions on both
sides of the circuit-breaker are suitable for Fig. 4 Configuration of CT and VT circuits
being connected. For this purpose the Syn-
chronism and Voltage Check function is En- 201 CT Starpoint:
abled in this example. In this application the CTs are connected as
138 Fault Locator: shown below in Figure 5. The polarity of the
Following fault clearance an inspection of the CT connection must be selected correctly to
fault location may be needed to check that ensure correct response by the protection.
there is no permanent damage or risk of fur- For this purpose the position of the starpoint
ther faults at the fault location. Particularly
on longer lines it is very helpful to have an
indication of the fault location to allow faster
access by the inspection team. For this pur-
pose the fault locator is Enabled in this ex-
ample.
140 Trip Circuit Supervision:
The monitoring done by the relay can be ex-
tended to include the trip circuit and trip
coils. For this purpose a small current is cir-
culated in the monitored circuits and routed
via binary inputs to indicate a failure. In this
example this function is not used and there-
fore set to Disabled. 19"
n 5. Masking I/O (configuration matrix) Fig. 5 Relay connections

The configuration matrix is used to route and al-
locate the information flow in the device. All the connection is indicated: In this example it
assignments of binary inputs and outputs, as well must be set towards line.
as LED’s, sequence of event records, user defined
logic, controls etc. are made in the matrix. 203 Rated Primary Voltage:
The VT ratio must be set correctly to ensure
n 6. User defined logic CFC accurate measured value output. It is also
If any special logic is required in the application, possible to set the protection parameters in
the CFC task can be used for this purpose. primary quantities. For correct conversion
from primary to secondary the VT and CT
n 7. Settings for power system data 1 data must be set correctly. In this application
7.1 Instrument transformers the VT primary voltage is 380 kV.
Under this heading power system parameters are 204 Rated Secondary Voltage (ph-ph):
applied. Place a ‘Tick’ in the box ‘Display addi- Set to 100 V as per VT data.
tional settings’ to include ‘advanced’ settings (des-
205 CT Rated Primary Current:
ignated by A, e.g. 0214A) in the displayed list.
Set to 1000 A as per CT data.

206 CT Rated Secondary Current: 214A Angle adjustment Usync-Uline:

Set to 1 A as per CT data. Note that this set- If there is a phase angle difference between
ting must correspond to the jumper settings the voltage Usync and Uline, for example due
on the measurement module (printed circuit to a power transformer with phase shifting
board). If this is not the case, the relay will vector group connected between the meas-
block and issue an alarm. Refer to device uring points, then this phase shift must be
manual for instructions on changing jumper set here. In this example the busbar is con-
settings. nected directly to the line so that there is
0° phase shift.
210 U4 voltage transformer is:
The 4th voltage measuring input may be used 215 Matching ratio Uline-Usync:
for a number of different functions. In this If the transformation ratio of the VT for
example it is connected to measure busbar line voltage and sync. voltage measurement
voltage for synchronising check (set Usync is not the same, then the difference must be
transformer) as shown in Figure 6. set here. In this application:
U prim Line 380
U sec Line
Ratio correction = = 01. = 105
.
U prim BUS 400
U sec BUS .
011
The required setting is therefore 1.05.

220 I4 current transformer is:
The 4th current measurement may be used
for a number of different functions. In this
case it is used to measure the Neutral Cur-
rent (of the protected line) by means of a
Holmgreen connection. See Figure 5.
221 Matching ratio I4/Iph for CT’s:
If the CT connected to I4 has a different ra-
tio, for example a core balance CT, to the
ratio of the CT measuring the phase cur-
rents of the protected circuit, this differ-
ence must be set here. In this application
the ratio is the same so the setting must be
Fig. 6 VT connections 1.00.
7.2 Power system data

211 Matching ratio Phase-VT to Open-delta-VT:
If in setting 210 the 4th voltage measuring in-
put is selected to measure the open-delta
voltage (3 U0) then this setting must be used
to configure the transformation ratio differ-
LSP2592.tif
ence between the phase VT and the open-

delta VT. As sync. check is applied this set-
ting has no relevance. Fig. 7 Power system data
212 VT connection for sync. voltage:
If the setting 210 for the 4th voltage measur- 207 System Starpoint is:
ing input is selected to measure the voltage The condition of the system starpoint
for sync. check this setting must be applied to earthing must be set here. If the system
define which voltage is used for sync. check. starpoint is not effectively earthed, isolated
In this example, the voltage connected to U4 or resonant earthed, then the distance pro-
is the phase-phase voltage L3-L1 as shown in tection response to simple earth faults will
Figure 6. be stabilised to prevent operation on tran-
sient earth fault currents. In this example
Solid Earthed applies.

230 Rated Frequency: When primary current flow is detected

Set the rated system frequency to 50 Hz or (measured current > pole open current:
60 Hz. Parameter 1130) then the trip command is
sealed in by the current flow and will only
235 Phase Sequence:
reset once the current flow is interrupted
The phase sequence of the system is usually
(refer to Figure 9). When the trip com-
positive, L1 L2 L3. If the system has nega-
mand is issued and no current flow is de-
tive phase sequence this can be set here. In
tected, the minimum trip command dura-
this example the phase sequence is positive
tion set here will apply. It must be set
(L1 L2 L3).
longer than the maximum time taken for
236 Distance measurement unit: the CB auxiliary contacts to open and in-
The distance measurement unit for the terrupt the current in the trip coil follow-
fault locator and certain line parameters ing the start of the trip command. The re-
can either be in km or in miles. In this ex- set conditions for the trip command can be
ample km is used. set with parameter “1135 RESET of Trip
Command”. From Table 2 the given circuit-
237 Setting format for zero seq. comp.:
breaker operating time is seen to be 60 ms.
The distance protection includes a zero
A safety margin of 50 ms is sensible so that
sequence compensation so that the same
a setting of 0.11 s is applied.
reach settings apply to phase and earth
faults. The zero sequence compensation 241A Maximum Close Command Duration:
can either be set as RE/RL and XE/XL pa- The close command must also have a mini-
rameters (standard format used by Siemens mum duration to ensure that the circuit-
in the past) or as the complex ratio KO by breaker can respond and that the auxiliary
means of a magnitude and angle setting. In contacts can interrupt the current flow
this example the setting will be applied as through the close coil. If, following a close
Zero seq. comp. factors RE/RL and XE/XL. command, a trip is issued due to switch on
to fault, the close command is reset imme-
7.3 Breaker diately by the new trip command. The
close command maximum duration
should be set at least as long as the maxi-
mum time required by the CB auxiliary
contact to interrupt the close coil current
after start of the close command. From Ta-
LSP2593.tif
ble 2 the given circuit-breaker operating

Fig. 8 Breaker parameters
time is seen to be 70 ms. A safety margin of
50 ms is sensible so that a setting of 0.12 s
is applied.
239 Closing (operating) time of CB:
This setting is only relevant if synchro 242 Dead Time for CB test-autoreclosure:
check with asynchronous switching is con- One of the test features in DIGSI is the CB
figured. Under asynchronous closing, the test-autoreclosure. For this test the circuit-
sync. check function will determine the in- breaker is tripped and reclosed under nor-
stant for issuing the close command so that mal load conditions. A successful test
the primary CB contacts close when the proves that the trip and close circuits and
switched voltages are in phase. For this the CB are in a fully functional state. As the
purpose the time that expires after applica- test causes a disruption to the power flow
tion of the close command to the close coil (either single phase or three phase), the
until the primary contacts of the CB make dead time should be as short as possible.
must be set here. From Table 2 the re- While a normal dead time must allow for
quired setting is 0.07 s. the time required by the fault arc to dissi-
pate (typically 0.5 s for three-pole trip and
240A Minimum TRIP Command Duration:
1s for single-pole trip), the test cycle must
The trip command to the circuit-breaker
only allow for the circuit-breaker mecha-
must have a minimum duration to ensure
nism to open and close. Here a dead time
that the CB responds and to prevent pre-
of 0.10 s is usually sufficient.
mature interruption of the current in the
trip coil which may cause damage to the
trip contact which is not rated to interrupt
such a large inductive current.

Fig. 9 Trip command seal in
n 8. Settings for Setting Group A

The setting blocks that are available in Setting
Group A depend on the selections made during
the selection of the device configuration (Heading
4). If the setting group changeover had been acti-
vated, a total of 4 setting groups would have been
available.
n 9. Settings for Power System Data 2

Further power system data, in addition to Power
LSP2594.tif
System Data 1, is set here. As these parameters are

inside Setting Group A, they can be modified be-
Fig. 10 Setting blocks available in Setting Group A for tween the setting groups if setting group change-
this application over is activated.
9.1 Power system

1103 Measurement: Full Scale Voltage (100 %):
For the indication and processing of meas-
ured values it is important to set the full
scale value on the primary side. This does
not have to correspond to the VT rated pri-
mary voltage. When the primary value cor-
responds to this setting the percentage
measured value will be 100 %. Other per-
centage measured values that also depend
LSP2595.tif
on voltage, such as for example power (P)

will also have the full scale indication de-
Fig. 11 Power system settings in Power System Data 2 pendant on this setting. In Table 2 the sys-
tem rated voltage is given and therefore set
at 400 kV.

1104 Measurement: Full Scale Current (100%):

For the indication and processing of mea-
sured values it is important to set the full
scale value on the primary side. This does
not have to correspond to the CT rated pri-
mary current. When the primary value cor-
responds to this setting the percentage
measured value will be 100 %. Other per-
centage measured values that also depend
on current, such as for example power (P)
will also have the full scale indication de-
pendant on this setting. In Table 2 the
rated apparent power of the line is stated at
600 MVA:
Rated MVA
Full scale current = Fig. 12 Polygon and line angle
3 ⋅ Full scale voltage
600
Full scale current = = 866 A
3 ⋅ 400 1110 x’-Line Reactance per length unit:
The measurement: Full scale current The line reactance per length unit (in this
(100 %) is therefore set to 866 A. example per km) is required for the fault
locator output in km (miles) and percent.
1105 Line Angle: In Table 2 this is given as 0.21 Ω/km pri-
The line angle setting is calculated from the mary. The setting can therefore be applied
positive sequence line impedance data. In as primary value 0.2100 /km, or it can be
this example: converted to a secondary value:
Z1 = 0.025 + j0.21 1000
X  CT ratio
Line angle = arctan  L  x'secondary = ⋅ x'primary = 1 ⋅ 0.21
VT ratio 380
 RL 
01
.
Line angle = 83° x'secondary = 0.0553
1211 Angle of inclination, distance charact.: The setting in secondary impedance is
This is usually set the same as the line an- 0.0553 /km.
gle. In this manner the resistance coverage
for all faults along the line is the same 1111 Line Length:
(Fig. 12). Therefore set for this application The line length setting in km (miles) is re-
the angle of inclination of the distance quired for the fault locator output. From
characteristic equal to the line angle which Table 2 set 80.0 km.
is 83° . 1116 Zero seq. comp. factor RE/RL for Z1:
1107 P,Q operational measured values sign: The zero sequence compensation setting is
The measured values P and Q are desig- applied so that the distance protection
nated as positive when the power flow is measures the distance to fault of all fault
into the protected object. If the opposite types based on the set positive sequence
sign is required, this setting must be reach. The setting is applied as RE/RL and
changed so that the sign of P and Q will be XE/XL setting; here RE/RL for Zone 1 with
reversed. In Table 2 the sign convention for the data for Line 1 from Table 2.
power flow states that exported power
RE 1  R0  1  013
. 
(flowing into the line) is designated as neg- = ⋅ − 1 = ⋅  − 1 = 14
.
ative. The setting here must therefore be R L 3  R 1  3  0.025 
reversed.
Apply setting RE/RL for Z1 equal to 1.40.

1117 Zero seq. comp. factor XE/XL for Z1:

The same consideration as for parameter
1116 above applies:
XE 1  X0  1  0.81 
= ⋅ − 1 = ⋅  − 1 = 0.95
X L 3  X 1  3  0.21 
Apply setting XE/XL for Z1 equal to 0.95.

1118 Zero seq. comp. factor RE/RL for Z1B…Z5:
As the overreaching zones cover the pro-
tected line as well as adjacent circuits, the
zero sequence compensating factor must
take the impedance parameters of the pro-
Fig. 13 Positive and zero sequence line impedance
tected line as well as the adjacent lines into profile
account. The Zone 2 reach has to be
co-ordinated with the protection on the
This is graphically shown for the X values
shortest adjacent feeder (Line 3) so that the
in Figure 13. A similar drawing can also be
Zone 2 reach will be used to determine this
made for the R values. Always use the Zone
setting. The other zone reaches are largely
2 setting in the X direction as a reference.
influenced by other system conditions such
Now apply the derived values, X21, R21,
as parallel and intermediate infeeds:
X20 and R20 to the following equation:
If the Zone 2 reach is set to 80 % of the to-
RE 1  R0  1  1376
. 
tal impedance up to the Zone 1 reach on = ⋅ − 1 = ⋅  − 1 = 1.38
Line 3 (shortest adjacent line) then the to- R L 3  R 1  3  2.672 
tal positive sequence impedance at the
Zone 2 reach limit is: Apply setting RE/RL for Z1B...Z5 equal to
1.38.
X 2 1 = 0.8 ⋅ (X Line1 + 0.8 ⋅ X Line3 )
1119 Zero seq. comp. factor XE/XL for Z1B…Z5:
X 2 1 = 0.8 ⋅ (80 ⋅ 0.21 + 0.8 ⋅17.5) = 24.64 This is the XE/XL setting corresponding to
the RE/RL setting 1118 above. Therefore
(X 2 − X Line1)
R 2 1 = R Line1 + ⋅ R Line3 apply the derived values, X21, R21, X20 and
X Line3 R20 to the following equation:
24.64 − 80 ⋅ 0.21
R 2 1 = 80 ⋅ 0.025 + ⋅15
. = 2.672 XE 1  X0  1  1034
. 
17.5 = ⋅ − 1 = ⋅  − 1 = 107
.
The corresponding zero sequence imped- X L 3  X 1  3 24.64 

ance is calculated as follows:
Apply setting XE/XL for Z1B...Z5 equal to
(X 2 − X Line1) 1.07.
X 2 0 = X0 Line1 + ⋅ X0 Line3
X Line3
24.64 − 80 ⋅ 0.21
X 2 0 = 80 ⋅ 0.81 + ⋅ 86.5
17.5
X 2 0 = 1034
.
(X 2 − X Line1)
R 2 0 = R0 Line1 + ⋅ R0 Line3
X Line3
24.64 − 80 ⋅ 0.21
R 2 0 = 80 ⋅ 013
. + ⋅ 7.5
17.5
R 2 0 = 1376
.

9.2 Line status

1130A Pole Open Current Threshold:
For a number of functions in the relay the
switching state of the circuit-breaker is an
important logical information input. This
can be derived via auxiliary contacts or by
measuring the current flow in the circuit.
With this parameter the current threshold
is set to determine the pole open condition
LSP2596.tif
of the circuit-breaker. If the phase current
measured by the relay is below this thresh-
old this condition for pole open detection Line status settings in Power System Data 2
is true.
In this application only the distance pro-
This setting should be as sensitive as possi-
tection will be used for switch on to fault
ble (setting equal to or lower than the
so that a setting of 0.05 s is sufficient.
smallest current pick-up threshold of a
protection function). Stray induced cur- 1134 Recognition of Line Closure with:
rents during a true open pole condition As stated above (1132A) the recognition of
may however not cause incorrect pick-up. line closure is important for the switch on
to fault protection functions. If the manual
In this example no special conditions have
close binary input is assigned in the matrix,
to be considered, so the default setting of
it will be one of the line closure detection
0.10 A is maintained.
criteria. If other circuit-breaker closing
1131A Pole Open Voltage Threshold: conditions such as auto-reclose or remote
As was described for the pole open current closing are applied then it is advisable to
above (1130A), the pole open voltage set- apply additional criteria for line closure de-
ting determines the threshold below which tection. In the table below the prerequisites
the voltage condition for pole open is true. for application of the individual conditions
are marked with X.
As single-pole tripping will be applied here
and the voltage transformers are located on 1134 Recognition of Manual Close BI al- CB aux allocated in VT on line side
the line side of the circuit-breaker, the set- Line Closure with: located in Matrix Matrix of CB
ting should be large enough to ensure that Manual Close BI X
the voltage induced on the open phase is
below this setting. Apply a setting that is at Voltage X
least 20 % below the minimum operating Current flow Always valid Always valid Always valid
phase to earth voltage. CB aux X
In this example the minimum operating Table 4 Prerequisites for application of individual conditions in Parameter 1134
voltage is 85 % of nominal voltage:
Setting In this example, the manual close binary
< 0.8 ⋅ 0.85 ⋅ 400 kV/380 kV ⋅ 100/ 3 < 41 input and CB aux contacts are not allo-
cated in the Matrix, so the conditions Volt-
Therefore apply a setting of 40 V. age and Current flow must be used for line
1132A Seal-in Time after ALL closures: closure detection. As the voltage trans-
When the feeder is energised the switch on formers are on the line side, the setting
to fault (SOTF) protection functions are Current or Voltage or Manual Close BI is
activated. The line closure detection condi- applied. Note that the inclusion of Manual
tions are set with parameter 1135 below. Close is of no consequence because the bi-
This seal-in time setting applies to all line nary input is not allocated in the Matrix.
closure detections other than the manual
close binary input condition. This direct
detection of circuit-breaker closing re-
sponds almost at the same instant as the
primary circuit-breaker contact closing. A
fairly short seal in time can therefore be set
here to allow for pick-up of the desired
protection functions.

1135 RESET of Trip Command: therefore:

The trip command duration must always
2 ⋅ 50 ⋅ 0.0179
be long enough to allow the circuit-breaker Rcable =
auxiliary contacts to interrupt the current 2.5
flowing through the trip coil. The most re- Rcable = 0.72 Ω
liable method for sealing in the trip com-
at 1 A nominal secondary current, this re-
mand is the detection of current flow in the
lates to:
primary circuit through the CB. The auxil-
iary contact status may be used as an addi- P' = Rcable ⋅ IN CT2 +Prelay
tional condition. This is helpful when trip P' = 0.72 ⋅ 12 + 0.1
commands are issued in the absence of pri- P' = 0.82 VA
mary current flow, e.g. during testing or by
From Table 2, the CT data is 5P20 20 VA,
protection functions that do not respond
therefore:
to current flow such as voltage or fre-
quency protection. In this example, the 20 + 3
n' = 20 ⋅ = 120
auxiliary contacts are not allocated in the 0.82 + 3
Matrix so that the trip command is reset
with Pole Open Current Threshold only. with this value, the setting can then be cal-
culated:
1140A CT Saturation Threshold: 120
CT saturation is normally detected by CT Sutaration Threshold = ⋅1 A = 24 A
5
monitoring of harmonic content in the The applied setting in this case is therefore
measured current. This is not possible for 24.0 A.
protection response below 1 cycle as at
least one cycle of recorded fault current is 1150A Seal-in Time after MANUAL closures:
required to determine the harmonic con- This setting is only applicable when the
tent. Below one cycle the CT saturation manual close binary input is allocated in
condition is therefore set when the current the Matrix (refer to setting 1134 above).
exceeds this threshold. The following cal- The time applied here should allow for the
culation gives an approximation of this circuit-breaker response and any addi-
current threshold: tional delays such as sync. check release
which can occur between the initiation of
n' the binary input and closure of the CB pri-
CT Sutaration Threshold = ⋅ IN
5 mary contacts. In this example, the manual
close binary input is not allocated so this
with
setting is of no consequence and therefore
PN + Pi left on the default value of 0.30 s.
n' = n ⋅ = actual overcurrent factor
P'+ Pi 1151 Manual CLOSE COMMAND generation:
P’ = the actual burden connected to the If the manual close binary input is allo-
secondary CT relay burden + CT cated, it may be used to generate a close
secondary connection cable burden command to the circuit-breaker in the re-
In this example only the 7SA relay is lay. Alternatively the input may be used
connected to the CT, so that the relay bur- only to inform the relay that a manual
den is 0.05 VA per phase. Due to the close has been issued externally to the cir-
Holmgreen connection, the maximum cuit-breaker. If the relay has to generate a
burden for earth currents is therefore twice close command following the initiation,
0.05 VA =0.1 VA. this can be done with or without sync.
The CT secondary cable connection bur- check if the internal sync. check function is
den is calculated as follows: available. In this example the manual close
binary input is not allocated so this setting
2 ⋅ l cable ⋅ ρCU should be set to NO.
Rcable =
acable
lcable = 50 m
ρCU = 0.0179 Ωmm2/m
acable = 2.5 mm2

1152 MANUAL Closure Impulse after n 10. Distance protection, general settings –
CONTROL: Setting Group A
If the internal control functions are used, 10.1 General
either via front keypad or system interface,
the issued control-CLOSE command to the
circuit-breaker can be used to activate the
protection functions in the same manner
as the manual-close binary input would.
The setting options provided consist of all
the configured controls in the device. In
this example, the internal control functions
are not used, so this setting is left on the
default value: <none>
LSP2598.tif
9.3 Trip 1/3-pole
As 1 and 3-pole tripping is applied in this
Fig. 16 General settings for distance protection
example, the following settings must be applied:
1201 Distance protection is:

If the distance protection must be switched
off in this setting group, it can be done so
LSP2597.tif
with this setting.

In this example, only one setting group is
Fig. 15 Trip 1/3-pole settings in Power System Data 2 used and the distance protection function
is required so this setting is left on the de-
1155 3 pole coupling: fault value which is ON.
When single-pole tripping is applied the
1202 Phase Current threshold for dist. meas.:
relay must select the faulted phase and trip
Although the distance protection responds
single pole selectively. In the event of 2 si-
to the impedance of the faulted loop, a
multaneous faults, e.g. inter-circuit fault
lower limit must apply for the current flow
on double circuit line, the relay detects two
before the distance protection responds. If
faulted phases, but only one of the two is
the system conditions can not ensure that
inside a tripping zone. To ensure single-
this minimum current flows during all in-
pole tripping under these conditions, set
ternal short-circuit faults, special weak
3-pole coupling to with Trip.
infeed measures may be required (see
1156A Trip type with 2phase faults: chapter 14). It is common practice to apply
Phase to phase faults without ground, can a very sensitive setting here so that the
be cleared by single-pole tripping. On cir- back-up functionality of the distance pro-
cuits where such faults occur frequently, tection for remote faults on other circuits
e.g. conductor clashing due to conductor is effective. The default setting of 10 % is
galloping with ice and wind conditions, therefore commonly applied.
single-pole tripping for 2-phase faults can In this application, no special conditions
improve availability of the circuit. The set- exist, so the default setting of 0.10 A is ap-
ting at both line ends must be the same. If plied.
single-pole tripping is selected, either the
1211 Angle of inclination, distance charact.:
leading or lagging phase will then be trip-
This setting was already discussed and ap-
ped at both line ends. In this example 2-
plied in chapter 9.1 “Power system”, where
phase faults will be tripped 3-pole.
its association with the line angle was de-
scribed. It is set to 83° .
1208 Series compensated line:
On feeders in the vicinity of series capaci-
tors, special measures are required for di-
rection measurement.
This application is without series capaci-
tors on the protected or adjacent feeders,
so the setting applied is NO.

1232 Instantaneous trip after SwitchOnToFault: this purpose, the smallest load impedance
When the protected circuit is switched off, and the largest load impedance angle must
a permanent fault (e.g. working earth or be determined (refer to Fig. 17).
broken conductor on ground) may be
The load encroachment area is set for
present. After switching on the circuit,
phase to earth loops (parameter 1241 and
such faults must be cleared as fast as possi-
1242) and for phase to phase loops (para-
ble. It is common practice to activate
meter 1243 and 1244) separately. Normally
non-selective stages with fast tripping for
load conditions will not cause earth fault
switch on to fault conditions. In the dis-
detection as no zero sequence current is
tance protection a number of alternatives
present in the load. In the event of single-
exists:
pole tripping of adjacent circuits, an earth-
Setting Distance protection during SOTF
fault detection and increased load current
flow may be present at the same time. For
Inactive No special measures such contingencies, the load encroachment
With pickup (non-directional) All distance zones are released for must also be set for earth-fault characteris-
instantaneous tripping tics.
With Zone Z1B The Zone Z1B is released for in- U operation min
stantaneous tripping and will oper- Rload min =
ate with its set direction if a polaris- 3 ⋅ I load max
ing voltage is available
Zone Z1B undirectional The Zone Z1B is released for in- From Table 1, the minimum operating
stantaneous tripping and will oper- voltage is 85 % of nominal system voltage,
ate as a non-directional zone.
(MHO characteristic as forward and
and the maximum load current is 250 % of
reverse zone) the full load apparent power.
Table 5 Setting alternatives for SOTF with distance U operation min = 0.85 ⋅ 400 kV = 340 kV
protection
600 MVA
I load max = 2.5 ⋅ = 2170 A
It is recommended to use the distance pro- 3 ⋅ 400 kV
tection for SOTF conditions. In many cases
By substituting these values in the above
the setting “with pickup (non-directional)”
equation:
would result in a reach that operates due to
heavy load inrush, e.g. when large machines 340 kV
and transformers are connected to the Rload min = = 90.5 Ω
3 ⋅ 2170
feeder so that the energising current is
more than twice the full load current. In To convert this to a secondary value, mul-
these cases the Zone Z1B can be applied as tiply it with the factor 0.2632 (Table 2) to
its reach is typically only between approx. obtain the setting 23.8 . As worst case
120 % and 200 % of the protected feeder. conditions are assumed, a safety factor is
Of special interest is the application of not required. If the parameters for calcula-
Zone Z1B undirectional. If the local busbar tion are less conservative, a safety factor,
can be energised from the remote end via e.g. 10 to 20 % may be included in the cal-
the protected feeder, then SOTF conditions culation.
for busbar faults can be provided by appli- 1242 PHI load, maximum Load Angle (ph-e):
cation of this setting. Note that the line clo- To determine the largest angle that the load
sure detection should not be with the impedance may assume, the largest angle
voltage condition in this case, as the live between operating voltage and load current
line voltage prior to energising the busbar must be determined. As load current ide-
would prevent the SOTF release. ally is in phase with the voltage, the differ-
In this example, the local bus will not be ence is indicated with the power factor
energised via the feeder so the setting with cos ϕ. The largest angle of the load impe-
Zone Z1B is applied. dance is therefore given by the worst,
1241 R load, minimum Load Impedance (ph-e): smallest power factor. From Table 2 the
The settings 1241 to 1244 determine the worst power factor under full load condi-
“load encroachment area” for the distance tions is 0.9:
relay setting characteristic. The distance ϕ load − max = arc cos(power factormin )
zone settings must exclude the load area in
the impedance plane so that operation is ϕ load − max = arc cos(0.9) = 26 °
only possible under fault conditions. For

The power factor under full load condi- This achieves fast tripping for all faults on
tions should be used for this calculation, as the feeder although some non-selective
under lightly loaded conditions the VAR trips can also occur. This is tolerated in
flow may dominate, but under these condi- such a scheme because following all fast
tions the load impedance is not close to the trips there is an automatic reclosure. In this
set impedance reach. In this case the setting example, the Z1B will only be controlled by
for PHI load, maximum Load Angle (ph-e) the teleprotection, the setting NO is there-
is 26° . fore applied.
1243 R load, minimum Load Impedance
10.2 Earth faults
(ph-ph):
No distinction is made in this example be-
tween the maximum load during phase to
earth pickup (adjacent circuit single-pole
open) and phase to phase pickup, e.g. when
parallel circuit is three-pole tripped. There-
fore the same setting as for 1241 is applied
LSP2599.tif
here, being 23.8 .
1244 PHI load, maximum Load Angle (ph-ph):
Fig. 18 Earth-fault settings for distance protection
Again the same setting as for the phase to
earth loop is applied here, being 26° .
1203 3I0 threshold for neutral current pickup:
The distance protection must identify the
faulted loop to ensure correct response. If
an earth-fault is present, this is detected by
the earth-fault detection. Only in this case
will the three earth loop measurements be
released subject to further phase selection
criteria. The earth current pickup is the
most important parameter for the earth-
fault detection. Its threshold must be set
below the smallest earth current expected
for faults on the protected feeder. As the
distance protection is also set to operate as
Fig. 17 Load encroachment characteristic backup protection for remote external
faults, this setting is set far more sensitive
1317A Single pole trip for faults in Z2: than required for internal faults. In chapter
For special applications, single-pole trip- 3 the minimum single-phase fault current
ping by Zone 2 can be applied. However, for internal faults neglecting fault resis-
time delayed protection stages are usually tance was calculated to be 1380 A.
applied with 3-pole tripping. To allow for fault resistance and reach into
In this example, 3-pole tripping in Zone 2 adjacent feeders for back-up, the setting
is desired so the default setting of NO is left applied here should be substantially lower
unchanged. than this calculated value. In this example,
the default value of 0.10 A secondary
1357 Z1B enabled before 1st AR (int. or ext.): (100 A primary) is maintained.
In this example a teleprotection scheme
(POTT) is applied. The controlled Zone
Z1B operation is therefore subject to the
signals from the teleprotection scheme.
In an application where no teleprotection
scheme is applied or in the case of a recep-
tion failure of the teleprotection scheme,
the Zone Z1B can also be controlled by the
auto-reclose function.

1204 3U0 threshold zero seq. voltage pickup: 1209A Criterion for earth fault recognition:
A further criteria for earth-fault detection For the settings 1203 and 1204 above, and
is the zero sequence voltage. In an earthed in Fig. 19, the method and logic of the
system, zero sequence voltage is always earth-fault detection were explained. With
present during earth faults and it decreases this setting the user has the means to influ-
as the distance between the measuring ence the earth-fault detection logic. In
point and the fault location increases. This earthed systems it is recommended to use
threshold setting is therefore also used for the very reliable OR combination of zero
earth-fault detection as shown in the logic sequence current and voltage for the earth-
diagram, Fig. 19. When the zero sequence fault detection. As mentioned before, these
source impedance is large, the zero se- two criteria supplement each other so that
quence current component in the fault small zero sequence current is often associ-
current may become small. In such an ated with large zero sequence voltage at
event, the zero sequence voltage will how- weak infeeds and the other way around at
ever be relatively large due to the small zero strong infeeds. The AND setting is only for
sequence current flowing through the large exceptional conditions when, for example,
zero sequence source impedance. the zero sequence voltage or current on
For secure earth-fault detection the default their own are not a reliable indicator for
setting of 5 V is maintained. If system un- earth faults.
balance during unfaulted conditions causes In this example, the default setting OR is
larger zero sequence voltages then this set- maintained for the reasons stated above.
ting should be increased to avoid earth-
fault detection under these circumstances.
Fig. 19 Earth-fault detection logic
1207A 3I0> pickup stabilisation (3I0>/Iph max):

In the event of large phase currents, the
system unbalance (e.g. non-transposed
lines) and CT errors (e.g. saturation) can
cause zero sequence current to flow via the
measuring circuit of the relay although no
earth fault is present. To avoid earth-fault
detection under these conditions, the zero
sequence current pickup is stabilised by
this set factor.
Unless extreme system unbalance or excep- Fig. 20 Stabilised zero sequence pickup threshold
tionally large CT errors are expected, the
default setting of 10 % i.e. 0.10 can be
maintained, as is done for this example.

1221A Loop selection with 2Ph-E faults:

If some fault resistance (arc voltage) is
present, then the measured fault loop im-
pedances are affected by this additional
voltage drop in the short-circuit loop. In
the case of 2Ph-E faults this is most severe
as the current in the fault resistance stems
from 3 different short-circuit loops. Theo-
retical analysis and simulations show the
following distribution of the measured
loop impedances for a 2Ph-E fault:
The influence of load (remote infeed and
load angle) can increase or decrease the ro-
tation of the measured fault resistances.
The leading phase to earth loop will how-
ever always tend to produce an overreach. Fig. 21 Impedance distribution for 2ph-E fault with fault resistance
For this reason, the default setting of block
leading Ph-E loop will be used in this ex-
ample. If the application is on a double cir-
cuit line where simultaneous earth faults
on both lines can occur, the setting only
phase-earth loops or all loops should be
used to avoid blocking of the internal fault
loop by this setting. Of course additional
grading margin must be applied for Zone 1
in this case to avoid an overreach during an
external 2Ph-E fault.
LSP2600.tif
10.3 Time delays
Fig. 22 Time delay setting for the distance zones
1210 Condition for zone timer start:
During internal faults, all time delayed
zones pickup unless there is substantial
fault resistance and very strong remote
infeed.
Although the fault in Fig. 23 is an internal
fault it is measured only by Zone Z4 due to
the fault resistance and strong remote in-
feed. If all zone timers are started by the
distance pickup, the fault will be cleared by
the relay with the set Zone 2 time after
fault inception because the measured im-
pedance moves into Zone 2 as soon as the
remote, strong infeed trips the breaker on
the right hand side.
From the timing diagram in Fig. 24 the in-
fluence of this setting can be seen. If the
zone timers are started with distance
pickup, the trip signal is issued with Zone 2
time delay (250 ms) after fault inception
(distance pickup) although the Zone 2 only
picks up some time later when the remote
end has opened the circuit-breaker on the
right hand side. The timing of the trip sig-
nals is therefore as if the fault had been in-
side the Zone 2 all along. For external fault
back-up tripping similar operation by
higher zones is achieved. This mode of op- Fig. 23 Influence of fault resistance and remote infeed on measured impedance
eration will be applied in this example, so
the setting with distance pickup is applied.

1306 T1 multi-ph, delay for multi phase faults:

Also here the zone 1 is usually operated as
fast tripping (instantaneous) underreach
zone. For the fastest tripping all Zone 1
times are set to 0.00 s. Refer also to setting
1305 above.
1315 T2-1phase, delay for single phase faults:
For the Zone 2 and higher zones the co-or-
dination time must be calculated. This
time must ensure that time graded tripping
remains selective.
In Fig. 25 the parameters that need to be
considered for the time grading margin are
shown. The values entered apply for this
Fig. 24 Timing diagram for fault in Fig. 23
example and correspond to the worst case
conditions. The required time grading
margin is therefore 250 ms. The Zone 2 is
graded with the Zone 1 at the remote feed-
ers so that a single time grading step is re-
quired (refer to Table 1). Set this time for
single-phase faults to 0.25 s. For special ap-
plications, the trip time of single-phase
faults may be set here to differ from that
for multi-phase faults which is set below
with parameter 1316.
As the Zone 2 will only trip three-pole in
this example, and no special consideration
is given to single-phase faults, this time is
Fig. 25 Time chart to determine time step for graded
protection
set the same as 1315 above to 0.25 s.
1325 T3 delay:
If co-ordination with other distance or From Table 1, the required time delay for
overcurrent protection is required, the set- this stage is two time steps = 0.50 s.
ting “with zone pickup” can be applied. 1335 T4 delay:
For the scenario described in Fig. 23 and 24 From Table 1, this stage is not required so
this will however result in additional time the time delay can be set to infinity, s.
delay (∆t in Fig. 24). During external faults
with backup protection operation this time 1345 T5 delay:
delay may become very long, often a full From Table 1, the required time delay for
time grading step so that reach and time this stage is three time steps = 0.75 s.
grading must be applied more conserva- 1355 T1B-1phase, delay for single ph. faults:
tively. The zone Z1B will be used for the
1305 T1-1phase, delay for single phase faults: teleprotection in a POTT scheme. For this
The Zone 1 is usually operated as fast trip- application no time delay is required so the
ping (instantaneous) underreach zone. For setting here is 0.00 s. For special applica-
the fastest tripping all Zone 1 times are set tions, the trip time of single-phase faults
to 0.00 s. For special applications, the trip may be set here to differ from that for
time of single-phase faults may be set here multi-phase faults which is set below with
to differ from that for multi-phase faults parameter 1356.
which is set below with parameter 1306.

1356 T1B-multi-ph, delay for multi ph. faults:

As stated above the Zone Z1B will be used
for the teleprotection in a POTT scheme.
For this application no time delay is re-
quired so the setting here is 0.00 s.
n 11. Distance zones (quadrilateral) – Setting

Group A
LSP2601.tif
11.1 Zone Z1
1301 Operating mode Z1: Fig. 26 Distance zone settings (Zone 1)
In the case of quadrilateral distance pro-
tection zones, the user may select the oper-
ating mode for each zone as either “for -
ward”, “reverse”, “non-directional” or
“inactive”. When the zone is “inactive”, it
does not produce any pickup signals or
trip. The other options can be seen in the
adjacent diagram where Z1, Z1B, Z2 and
Z4 are set in the forward direction. Z3 is set
in the reverse direction and Z5 is set non-
directional. In this example, Zone 1 must
be set in the forward direction.
1302 R(Z1), Resistance for ph-ph faults:
As the distance protection is applied with
polygonal (quadrilateral) tripping charac-
teristics, the zone limits are entered as re-
sistance (R) and reactance (X) settings. A
separate resistance reach setting is available
for ph-ph measured loops and ph-e meas-
ured loops. This setting is for the ph-ph
Fig. 27 Quadrilateral zone diagram
loops. With setting “1211 Angle of inclina-
tion, distance charact.” the polygon
R-reach is inclined such that it is parallel to To obtain the largest value of Rarc , which is
the line impedance (refer to Figure 12). required for the setting, the smallest value
The resistance settings of the individual of fault current must be used (calculated in
zones therefore only have to cover the fault Chapter 3):
resistance at the fault location. For the 2500 V ⋅ 2 ⋅ 5 m
Zone 1 setting only arc faults will be con- Rarc = = 12.7 Ω
1967 A
sidered. For this purpose the arc resistance
will be calculated with the following equa- By addition of a 20 % safety margin and
tion: conversion to secondary impedance (factor
U arc from Table 2) the following minimum set-
Rarc = ting is calculated (division by 2 because Rarc
IF appears in the loop measurement while the
The arc voltage (Uarc) will be calculated us- setting is done as phase impedance or posi-
ing the following rule of thumb which pro- tive sequence impedance):
vides a very conservative estimate (the 1.2 ⋅12.7 ⋅ 0.2632
estimated Rarc is larger than the actual R(Z1) = = 2.01 Ω (sec.)
2
value):
This calculated value corresponds to the
U arc = 2500 V ⋅ l arc whereby larc is the smallest setting required to obtain the de-
length of the arc. sired arc resistance coverage. Depending
The length of the arc is greater than the on the X(Z1) reach calculated (see next
spacing between the conductors (ph-ph), page), this setting may be increased to ob-
because the arc is blown into a curve due to tain the desired Zone 1 polygon symmetry.
thermal and magnetic forces. For estima-
tion purposes it is assumed that larc is twice
the conductor spacing.

Therefore, looking ahead at the setting re- To compensate for this influence, the max-
sult for “1303 X(Z1), Reactance” below, we imum value (for practical purposes) of the
see that 3.537 Ohm are applied. For over- ratio of I2/I1 is required. This is given in
head line protection applications, the fol- Table 2 as the value 3. The maximum
lowing rule of thumb may be used for the tower footing resistance that is measured
R(Z1) setting: by the relay in the fault loop is therefore:
0.8 ⋅ X(Z1) < R(Z1) < 2.5 ⋅ X(Z1)  I2
RTF = 1 +  ⋅ effective tower footing R
In this example the lower limit applies, so  I1 
the setting for R(Z1) is:
RTF = (1 + 3) ⋅ 1.5 = 6 Ω (prim.)
R(Z1) = 0.8 ⋅ 3.537 = 2.830 Ω (sec.)
The arc voltage for the earth faults is calcu-
This setting is then applied, 2.830 . lated as follows using the conductor to
1303 X(Z1), Reactance: tower/ground spacing given in Table 2:
The reactance reach is calculated based on Uarc = 2500 V ⋅ larc
the distance reach that this zone must pro-
vide. In Table 1 the reach of Zone 1 is spec- Uarc = 2500 V ⋅ 2 ⋅ 3 m = 15 kV
ified as 80 % of Line 1. Therefore: To obtain the largest value of Rarc , which is
X(Z1) = 0.8 ⋅ XLine 1 required for the setting, the smallest value
of fault current must be used (calculated in
X(Z1) = 0.8 ⋅ 80 ⋅ 0.021 = 13.44 Ω (prim.) Chapter 3):
This is converted to a secondary value by 15 kV
multiplying with the conversion factor in Rarc = = 10.9 Ω
1380 Ω
Table 2:
The total resistance that must be covered
X(Z1) = 13.44 ⋅ 0.2632 = 3.537 Ω (s)
during earth faults is the sum of Rarc and
This setting is then applied, 3.537 . RTF . A safety factor of 20 % is included and
the result is converted to secondary values
1304 RE(Z1), Resistance for ph-e faults:
(division by factor (1 + RE/RL), because
The phase to earth fault resistance reach is
Rarc and RTF appear in the loop measure-
calculated along the same lines as the “1302
ment while the setting is done as phase im-
R(Z1)” setting for ph-ph faults. For the
pedance or positive sequence impedance):
earth fault however, not only the arc volt-
age must be considered, but also the tower 1.2 ⋅ (10.9 + 6) ⋅ 0.2632
RE(Z 1) = = 2.22 Ω(sec.)
footing resistance. From the graph in Fig- (1 + 14
.)
ure 29 it is apparent that although the indi-
vidual tower footing resistance is 15 Ω
(Table 2) the resultant value due to the This calculated value corresponds to the
parallel connection of multiple tower foot- smallest setting required to obtain the de-
ing resistances is less than 1.5 Ω. sired resistance coverage. Depending on
the X(Z1) reach calculated above, this set-
ting may be increased to obtain desired
Zone 1 polygon symmetry. The setting re-
sult for “1303 X(Z1), Reactance” is
3.537 Ω. For overhead line protection ap-
plications, the following rule of thumb may
be used for the RE(Z1) setting; note that
the lower limit is the same as for ph-ph
faults – this ensures fast Zone 1 tripping,
while the upper limit is based on the loop
Fig. 28 Combination of arc voltage and tower footing reach – this avoids overreach:
resistance
From Fig. 28 it can be seen that the remote

infeed (I2) will introduce an additional
voltage drop across the “effective tower
footing resistance” which will be measured
in the fault loop by the relay (this effect is
also shown in Figure 23).

Fig. 29 Effective tower footing resistance
XE
1+
0.8 ⋅ X(Z 1) < RE(Z 1) < XL ⋅ 2.5 ⋅ X(Z 1)
RE
1+
RL
In this example the lower limit applies, so
the setting for RE(Z1) is:
RE(Z 1) = 0.8 ⋅ 3537
. = 2.83 Ω(sec.)
This setting is then applied, 2.83 .


The Zone 1 is required to trip as fast as
possible, therefore this time is set to 0.00 s.
The Zone 1 is required to trip as fast as
possible, therefore this time is set to 0.00 s.
1307 Zone Reduction Angle (load compensa-
tion):
The Zone 1 may under no circumstances
operate for external faults as this would
mean a loss of selectivity. The influence of
remote infeed in conjunction with fault re-
sistance must be considered in this regard.
From the voltage and current phasors in Fig. 31 Curves for selection of Alpha angle setting
Fig. 30 the influence of the transmission
angle (TA), i.e. angle between the voltages From Table 2 the maximum “Transmis -
VA and VB, on the measured fault resis- sion Angle “ for this application is given as
tance can be seen. In the impedance plane 35° . If this is checked in Figure 31 together
the phasor I2/I1 RF is rotated downwards by with the R1/X1 setting value of 0.8 (see
the transmission angle. The risk of the ex- above 2.830/3.537 = 0.8), then the required
ternal fault encroaching into Zone 1 is “Alpha Angle” setting is less than 15° (by
shown. To prevent this, the Zone 1 X set- using the TA = 40 curve). A setting of 15°
ting characteristic is tilted downwards by is therefore applied.
the “Alpha angle”. A detailed calculation of
the “Alpha angle” is complicated and
heavily dependant on changing system
conditions. A worst practical case is there-
fore selected to fix the “Alpha angle” set-
ting. For this purpose the largest
transmission angle that can occur in the
system during normal overload conditions
must be applied to the following set of
curves.
Fig. 30 Transmission angle for Alpha angle setting

11.2 Zone Z1B
1351 Operating mode Z1B (overreach zone):

The Zone Z1B will be used for the
teleprotection scheme POTT in this appli-
cation. For this the Zone Z1B must be ap-
plied as forward overreach zone.
LSP2602.tif
1352 R(Z1B), Resistance for ph-ph faults:
As was the case for the Zone 1 settings, this
setting must cover all internal arc faults. Fig. 32 Settings for Zone Z1B
The minimum setting therefore equals the
R(Z1) setting, 2.830 Ω. However, addi- 1354 RE(Z1B), Resistance for ph-e faults:
tional reach is set for the Z1B compared to As was the case for the Zone 1 settings, this
Z1, because this is an overreach zone while setting must cover all internal arc faults.
Z1 is set to underreach. The amount of ad- The minimum setting therefore equals the
ditional R reach mainly depends on the ra- RE(Z1) setting, 2.83 Ohm. As described
tio of R reach to X reach setting. For the for setting “1352 R(Z1B)” additional reach
Zone Z1B the following limit is recom- is usually applied and the following rule of
mended: thumb is used for the RE(Z1B) setting:
X(Z1B) < R(Z1) < 4 ⋅ X(Z1) XE XE
1+ 1+
Looking ahead at the applied setting for XL ⋅ X(Z1B) < RE(Z1B) < XL ⋅ 4 ⋅ X(Z1B)
RE RE
“1353 X(Z1B), Reactance” which is 6.633 1+ 1+
Ohm, it is apparent that the lower limit will RL RL
apply. The setting of R(Z1B) therefore also In this example the lower limit applies, so
is 6.633 . the setting for RE(Z1B) is:
1353 X(Z1B), Reactance: (1 + 1.07)

RE(Z 1B) = ⋅ 6.633 = 5.769 Ω(sec.)
The Zone Z1B must be set to overreach (1 + 1.38)
Line 1. The minimum setting is 120 % of
the line reactance. In practice, however, a
setting of 150 % or greater is applied unless 1355 T1B-1phase, delay for single ph. faults:
the line in question is extremely long. In this POTT scheme both single and
multi-phase faults must be tripped without
The risk of underreach due to the effects
additional delay. The setting 0.00 s is there-
shown in Figs. 21, 23 and 30 are then
fore applied.
avoided. For this application, medium
length line, a reach of 150 % is selected: 1356 T1B-multi-ph, delay for multi ph. faults:
In this POTT scheme both single and
X(Z1B) = 1.5 ⋅ XLine 1
multi-phase faults must be tripped without
X(Z1B) = 1.5 ⋅ 80 ⋅ 0.21 = 25.2 Ω (prim.) additional delay. The setting 0.00 s is there-
fore applied.
The applied setting therefore is:
1357 Z1B enabled before 1st AR (int. or ext.):
X(Z1B) = 25.2 Ω(prim.) ⋅ 0.2632
This setting was already applied in Chapter
X(Z1B) = 6.633 Ω(sec.) 10.1 and is repeated here with the other
Z1B settings. The setting is NO.
6.633 .

11.3 Zone Z2 6.485

RE(Z2) = ⋅ 2.83 ⋅1.2 = 4.98 Ω(sec.)
80 ⋅ 0.21 ⋅ 0.2632
10.3 and is shown here again with all the
Zone 2 settings. The setting 0.25 s is ap-
plied.
LSP2603.tif
1356 T2-multi-ph, delay for multi phase faults:

Fig. 33 Settings for Zone Z2 10.3 and is shown here again with all the
1311 Operating mode Z2: plied.
The Zone Z2 is used as the first overreach-
ing time graded zone. It must therefore be 1317A Single pole trip for faults in Z2:
applied in the forward direction. This setting was already applied in Chapter
10.1 and is shown here again with all the
1312 R(Z2), Resistance for ph-ph faults: Zone 2 settings. The setting NO is applied.
Resistance coverage for all arc faults up to
the set reach (refer to Table 1) must be ap- 11.4 Zone Z3
plied. As this zone is applied with over-
reach, an additional safety margin is in-
cluded, based on a minimum setting equiv-
alent to the X(Z2) setting and arc resis-
tance setting for internal faults, R(Z1)
LSP2604.tif
setting. Looking ahead, X(Z2) is set to

6.485 . Therefore:
X(Z2) Fig. 34 Settings for Zone Z3
R(Z2)min = ⋅ R(Z 1)
X Line1(s)
1321 Operating mode Z3:
6.485 Zone Z3 is used as reverse time delayed
R(Z2)min = ⋅ 2.83 = 415
. Ω(sec.) back-up stage (refer to Table 1). It must
80 ⋅ 0.21 ⋅ 0.2632
therefore be set in the reverse direction.
The setting of R(Z2) therefore is 4.150 .
1313 X(Z2), Reactance:
Resistance coverage settings for backup
According to the grading requirement in
protection with distance protection zones
Table 1:
is defined by a lower limit and upper limit.
CTratio The lower limit is the minimum fault resis-
X(Z 2) = 0.8 ⋅ (X Line1 + 0.8 ⋅ X Line3 ) ⋅
VTratio tance (arc resistance) that must be covered,
and the upper limit is based on the corre-
X(Z 2) = 0.8(80 ⋅ 0.21 + 0.8 ⋅17.5) ⋅ 0.2632 sponding X reach setting. Note that for
X(Z 2) = 6.485 Ω(sec.) high resistance faults (not arc faults), the
other infeeds to the reverse fault cause se-
The applied setting therefore is 6.485 . vere underreach.
As no detailed values are available, it is safe
1314 RE(Z2), Resistance for ph-e faults:
to assume that the required arc resistance
Similar to the R(Z2) setting, the minimum
coverage is the same as that calculated for
required reach for this setting is based on
faults on Line 1. Therefore the setting for
the RE(Z1) setting which covers all internal
R(Z1), 2.830 Ω, defines the lower limit.
fault resistance and the X(Z2) setting
The upper limit is given by reach symmetry
which determines the amount of over-
constraints and states that R(Z3)<6 times
reach. Alternatively, the RE(Z2) reach can
X(Z3). Looking ahead, X(Z3) is set to
be calculated from the R(Z2) reach with
2.211 Ω, so the upper limit is 13.266 Ω.
the following equation:
A setting halfway between these limits is a
X(Z 2) safe compromise:
RE(Z 2) = ⋅ RE(Z 1) ⋅1.2
X Line 1(sec.)

R(Z 1) + 6 ⋅ X(Z 3) 11.6 Zone Z5

R(Z 3) =
2
2.83 + 6 ⋅ 2.211
R(Z 3) = = 8.048 Ω(sec.)
2
The applied setting therefore is 8.048 .
1323 X(Z3), Reactance:
According to the grading requirement in
LSP2606.tif
Table 1:
CTratio
X(Z3) = 0.5 ⋅ X Line 1 ⋅ Fig. 36 Settings for Zone Z5
VTratio
1341 Operating mode Z5:
X(Z3) = 0.5 ⋅ 80 ⋅ 0.211 ⋅ 0.2632 = 2.211 Ω(sec.) Zone Z5 is used as non-directional final
backup stage (refer to Table 1). It must
therefore be set as non-directional zone.
1324 RE(Z3), Resistance for ph-e faults: Resistance coverage settings for backup
Similar to the R(Z3) setting, the upper and protection with distance protection zones
lower limits are defined by minimum re- is defined by a lower limit and upper limit.
quired reach and symmetry. In this appli- The lower limit is the minimum fault resis-
cation set the RE(Z3) reach the same as tance (arc resistance) that must be covered,
R(Z3) to 8.048 . and the upper limit is based on the corre-
1325 T3 delay: sponding X reach setting. Note that for
This setting was already applied in Chapter high resistance faults (not arc faults), the
10.3 and is shown here again with all the other infeeds to the reverse fault cause se-
Zone 3 settings. The setting 0.50 s is ap- vere underreach.
plied. As no detailed values are available, the re-
quired arc resistance coverage is calculated
11.5 Zone Z4 for the arc voltage (5 m) and 50 % of nom-
inal current or 500 A primary.
2500 V / m ⋅ 2 ⋅ 5 m CTratio
R(Z 5)min = ⋅
500 A VTratio
2500 ⋅ 2 ⋅ 5
R(Z 5)min = ⋅ 0.2632 = 1316
. Ω(sec.)
LSP2605.tif
500
Fig. 35 Settings for Zone Z4

This setting would ensure detection in
Zone 5, if the arc voltage, as calculated in
1331 Operating mode Z4: Chapter 11.1, is for 5 m conductor spacing
Zone Z4 is not applied (refer to Table 1). It and the fault current is at least 500 A. The
must therefore be set as inactive direction. upper limit is given by reach symmetry
Further settings in this block are of no con- constraints and states that R(Z5)< 6 times
sequence and therefore not discussed here. X(Z5) + or X(Z5) –. Looking ahead, X(Z5)
is set to 17.782 Ω, so the upper limit is
106.69 Ω. This is far into the load area (re-
fer to parameter 1241, calculated in Chap-
ter 10.1). A setting of double the minimum
is a safe compromise:
R(Z5) = R(Z5)min ⋅ 2
R(Z5) = 13.16 Ω ⋅ 2 = 26.32 Ω

1343 X(Z5)+, Reactance for Forward direction: 2006 Power swing trip:
According to the grading requirement in When the power swing is severe and an out
Table 1: of step condition is reached, selective
power swing tripping should be applied in
CTratio
X(Z 5) = 1.2 ⋅ (X Line1 + X Line2) ⋅ the system to obtain stable islanded subsys-
VTratio tems. This relay is not positioned on such
X(Z 5) = 1.2 ⋅ (80 ⋅ 0.21 + 39.5) ⋅ 0.2632 an interconnection so out of step tripping
is not required, therefore setting is NO.
X(Z5) = 17.782 Ω (sec.)
2007 Trip delay after Power Swing Blocking:
The applied setting therefore is 17.782 . If during a power swing, which is detected
1344 RE(Z5), Resistance for ph-e faults: by the relay, an external switching opera-
Similar to the R(Z5) setting, the upper and tion takes place, a jump of the measured
lower limits are defined by minimum re- “swing impedance” takes place. This jump
quired reach and symmetry. In this appli- can reset the power swing detection. To
cation set RE(Z5) reach to the same as prevent tripping, if this impedance is inside
R(Z5). The applied setting therefore is the protected zones, a delay time of 0.08 s
26.32 . is set to allow the power swing measure-
ment to securely pick up again.
1345 T5 delay:
This setting was already applied in Chapter n 13. Teleprotection for distance protection –
10.3 and is shown here again with all the Setting Group A
plied.
1346 X(Z5)-, Reactance for Reverse direction:
For non-directional Zone Z5 the following
symmetry requirement can be applied, if
LSP2608.tif
no other conditions are specified:

0.5 ⋅ X(Z5)+ < X(Z5)- < 2 ⋅ X(Z5)+ Fig. 38 Teleprotection for distance protection settings
In this case the lower limit will apply so
that: 2101 Teleprotection for Distance prot. is:
In this application the teleprotection is
X (Z5)- = 0.5 ⋅ X (Z5)+
ON.
X (Z5)- = 0.5 ⋅ 17.782
X (Z5)- = 8.891 Ω(sec.) 2102 Type of Line:
The line is a two terminal line.
2103A Time for send signal prolongation:
n 12. Power swing – Setting Group A In the event of sequential operation at the
two line ends or very slow operation at one
end, the end that trips first may reset and
stop transmitting the send signal before the
slow end is ready to operate. The send sig-
nal prolongation time ensures that the trip
signal only resets once the remote end has
LSP2607.tif
had sufficient time to pickup.
Fig. 37 Power swing settings
2002 Power Swing Operating mode:

In the event of a power swing, tripping by
the distance protection due to the meas-
urement of the “swing impedance” must be
prevented. The setting all zones blocked is
therefore applied.

In Figure 39 the delayed pickup of the re-

mote end, after opening of the local cir-
cuit-breaker, must be considered to ensure
that the send signal is not reset too soon.
Channel timing is neglected as it adds on to
the safety margin. The times indicated in
the drawing apply in this example so that a
setting of 0.05 s is applied.
2109A Transient Block.: Duration external flt:
The transient blocking, also referred to as
current reversal guard, is required with the
POTT scheme, if parallel circuits are pre-
sent. During clearance of a fault on the
parallel circuit, the fault current may re- Fig. 39 Time chart to send signal prolongation time
verse on the protected circuit. To avoid
operation of the POTT scheme under these
conditions, the transient blocking times are
applied. To ensure that transient blocking
is only activated by external faults, it only
starts following reverse fault detection for
this time which is set to 80 % of the fastest
fault clearance on the parallel circuit (in-
LSP2609.tif
cluding CB operation).
Tw = 0.8 ⋅ (Tcircuit_ breaker + Trelay_ parallel_ Line − Trelay_ protected_ Line ) Fig. 40 Weak infeed settings
Tw = 0.8 ⋅ (60 ms + 10 ms − 20 ms) = 40 ms

end in addition to sending the echo. In this
application both Echo and Trip will be
The setting 0.04 s is applied. used.
2110A Transient Block.: Blk. T. after ext. flt: 2502A Trip /Echo Delay after carrier receipt:
Following clearance of the external fault, As the communication channel may pro-
the transient blocking condition must be duce a spurious signal (unwanted recep-
maintained until both line ends securely tion), a small delay is included for security
detect the new fault condition. For this purposes. Only if the receive signal is pres-
purpose the relay pickup time (re-orienta- ent for this time will the weak infeed func-
tion) and channel delay time must be con- tion respond. In the event of 3-pole open
sidered. condition of the circuit-breaker, this time
delay is bypassed and the echo is sent im-
Tb = 1.2 ⋅ (Tchannel_ delay + Trelay_re -orientation ) mediately. The default setting of 0.04 s is
Tb = 1.2(20 ms + 20 ms) = 48 ms appropriate for this application.
The setting 0.05 s is applied. 2503A Trip Extention / Echo Impulse time:
To ensure that the echo signal can be se-
n 14. Weak infeed (trip and/or echo) – curely transmitted, it must have a defined
Setting Group A minimum length. On the other hand, a
permanent echo signal is not desired.
2501 Weak Infeed function is:
Therefore the echo is sent as a pulse with
When the POTT teleprotection scheme is
this set length. If tripping is also applied,
used, the weak infeed function can be ap-
then this time also defines the length of the
plied for fast fault clearance at both line
internal trip signal (refer also to parameter
ends even if one line end has very weak or
240 A in Chapter 7.3). The default setting
no infeed. The weak infeed function must
of 0.05 s is appropriate for this application.
be activated at the line end where a weak
infeed can occur. If a strong infeed is as-
sured at all times, this function can be
switched off. The function may also be
used to only send an echo back to the
strong infeed, so that it can trip with the
POTT scheme, or to trip at the weak infeed

2504A Echo Block Time: 2601 Operating mode:

If weak infeed echo is applied at both line The distance protection is more selective
ends, then it must be avoided that a re- and more sensitive than the overcurrent
ceived echo signal is again sent as echo in a protection. The overcurrent protection is
continuous stream of echo signals. For this therefore only required when the distance
purpose, this blocking time is set to pre- protection is blocked due to failure of the
vent a new weak infeed operation before it voltage measuring circuit (emergency
expires. A secure setting of this timer is the mode). Therefore the operating mode is set
time required for a signal to be transmitted to ON: only active with loss of VT sec. cir.
from one end to the other and back again
2680 Trip time delay after SOTF:
(twice the channel delay time) plus a safety
Following line closure detection the
margin of 10 ms. In this application, a
“Switch On To Fault” function is activated
worst channel delay time of 20 ms is as-
(refer to parameters 1132A and 1134 in
sumed so that a setting of 0.05 s is appro-
Chapter 9.2). The backup overcurrent sta-
priate.
ges may also be used for SOTF tripping.
2505 Undervoltage (ph-e): This timer sets the time delay for SOTF trip
The weak infeed trip signal (phase selec- with a backup overcurrent. In this applica-
tive) is supervised by this undervoltage tion SOTF with backup overcurrent is not
threshold. At the weak infeed end the applied so the setting of this timer is not
source impedance is very large so that very relevant; leave the default value of 0.00 s.
small voltages are measured during faults
on the protected circuit. If this threshold is 15.2 I>> stage
set well below the minimum operational
ph-e voltage then weak infeed tripping is
secure and phase selective. Simulations and
practical experience have shown that a set-
ting of 50 % nominal ph-e voltage provides
good results. The default setting of 25 V is
LSP2611.tif
therefore applied.
2509 Echo Logic: Dis and EF on common chan- Fig. 42 I>> stage settings, backup overcurrent
nel:
If the distance protection and directional 2610 Iph>> Pickup:
earth-fault protection are both applied This high set stage is required to trip with
with teleprotection (Distance with POTT single time step grading. It should there-
and EF with directional comparison) the fore have a reach equivalent to the Zone 2
signals can be routed via one common setting. The setting should therefore be
channel or via two separate channels in the equal to the maximum 3-phase fault cur-
communication system. The weak infeed rent for a fault at the Zone 2 reach setting.
and echo logic must be set accordingly to
ensure correct response. In this applica- Based on the source and line impedances,
tion, two separate channels are used so the the following maximum fault current level
setting is NO. can be calculated for faults at the Zone 2
reach limit:
n 15. Backup overcurrent – Setting Group A U source
15.1 General 3 ⋅ Z tot
Ztot = sum of minimum positive sequence

source and line impedance up to Zone 2
reach (as only current magnitudes are be-
ing calculated, only the magnitude of the
LSP2610.tif
impedance is relevant)
Fig. 41 General settings, backup overcurrent

| Z tot | =|(Rsource_ min + 0.8 ⋅ (RLine1 + 0.8 ⋅ RLine2)) + j(X source_ min + 0.8 ⋅ (X Line1 + 0.8 ⋅ X Line2))|
| Z tot | =|(1 + 0.8 ⋅ (80 ⋅ 0.025 + 0.8 + 15
. )) + j(10 + 0.8 ⋅ (80 ⋅ 0.21 + 0.8 ⋅17.5))|
| Z tot | = 356
. + j34.64
| Z tot | = 34.8
The maximum three-phase fault current at

the Zone 2 reach limit therefore is:
400 kV
I 3ph Z2max = = 6636 A
3 ⋅ 34.8
As a secondary value, the setting applied for
I>> is therefore 6.64 A.
2611 T Iph>> Time delay:
This high set stage is required to trip with
single time step grading. Therefore set 0.25 s
which is one time step (refer to Fig. 25).
2612 3I0>> Pickup:
This high set stage is required to trip earth
faults with single time step grading. It
should therefore have a reach equivalent to
the Zone 2 setting. The setting should there-
fore be equal to the maximum single-phase
fault current for a fault at the Zone 2 reach
setting.
Based on the source and line impedances,
the following maximum fault current level
can be calculated for faults at the Zone 2
reach limit:
U source
3 ⋅ Z tot
Ztot = 1/3 of the sum of minimum positive,

negative and zero sequence source and line
impedance up to Zone 2 reach (as only cur-
rent magnitudes are being calculated, only
the magnitude of the impedance is relevant)
For the 3-phase fault level used in setting
2610, the total positive sequence impedance
was calculated. As the negative sequence
impedance equals the positive sequence
value, the Ztot for this setting can be calcu-
lated as follows:
|2 ⋅ Z tot _ 2610 + (R0 source_ min + 0.8 ⋅ (R0 Line1 + 0.8 ⋅ R0 Line2)) + j(X0 source_ min + 0.8 ⋅ (X0 Line1 + 0.8 ⋅ X0 Line2))|
|Z tot | =
3
. + 2.5 + 0.8 ⋅ (80 ⋅ 013
|(712 . + 0.8 ⋅ 7.5)) + j(69.28 + 20 + 0.8 ⋅ (80 ⋅ 0.81 + 0.8 ⋅ 86.5))|
|Z tot | =
3
|Z tot| = |7.58 + j65.49|
|Z tot| = 65.9

The maximum single-phase fault current at As a secondary value, the setting applied
the Zone 2 reach limit therefore is: for I> is therefore 1.74 A.
400 kV 2621 T Iph> Time delay:
I 3ph Z2max = = 3504 A
3 ⋅ 65.9 This stage is required to trip with the same
time delay as Zone 5, three time grading
As a secondary value, the setting applied steps. Therefore set 0.75 s which is three
for 3I0>> is therefore 3.50 A. time steps (refer to Fig. 25).
2613 T 3I0>> Time delay: 2622 3I0> Pickup:
This high set stage is required to trip with This stage is required to trip with time de-
single time step grading. Therefore set 0.25 s lay equal to Zone 5. It should detect earth
which is one time step (refer to Fig. 25). faults with similar sensitivity as Zone 5.
2614 Instantaneous trip via Teleprot./BI: Therefore, with the weakest infeed accord-
The backup overcurrent is only active ing to Table 2, an earth fault at the X reach
when the distance protection is blocked limit of Zone 5 will have the following cur-
due to failure of the secondary VT circuit rent magnitude:
(refer to setting 2601 in Chapter 15.1). If U nom_ sec
under these conditions a teleprotection sig- 3I 0 Z5_ min =
 XE 
nal is received from the remote end, the 3 ⋅ (X source_ max + X Z5_ sett ) ⋅ 1 + 
 XL 
tripping of the overcurrent protection may
be accelerated. This may be safely applied 100
3I 0 Z5_ min =
for this stage, because its reach is less than 3 ⋅ (100 ⋅ 0.2632 + 17.782) ⋅ (1 + 1.38)
the set Z1B. Therefore apply the setting 3I 0 Z5_ min = 0.55 A
YES. Note that for this function to work,
the binary input function “7110 >O/C As a secondary value, the setting applied
InstTRIP” must be assigned in parallel to for 3I0> is therefore 0.55 A.
the teleprotection receive binary input of 2623 T 3I0>> Time delay:
the distance protection. This high set stage is required to trip with
2615 Instantaneous trip after SwitchOnToFault: three time grading steps. Therefore set
This function is not applied (refer to set- 0.75 s which is three time steps (refer to
ting 2680 in Chapter 15.1). Therefore NO Fig. 25).
is set. 2624 Instantaneous trip via via Teleprot./BI:
The I>> stage is applied for this purpose,
15.3 I> stage refer to setting 2614 in Chapter 15.2.
Therefore set NO for this stage.
2625 Instantaneous trip after SwitchOnToFault:
This function is not applied (refer to set-
ting 2680 in Chapter 15.1). Therefore NO
is set.
LSP2612.tif
Fig. 43 I> stage settings, backup overcurrent
2620 Iph> Pickup:

This stage is required to trip with time de-
lay equal to Zone 5. It may not pick up due
to load (permissible overload). The per-
missible overload is twice the full load,
therefore:
2 ⋅ Rated MVA
I ph > Pickup ≥
3 ⋅ Full scale voltage
2 ⋅ 600
I ph > Pickup ≥ = 1732 A
3 ⋅ 400

15.4 Inverse stage 2670 Instantaneous trip via Teleprot./BI:

The I>> stage is applied for this purpose,
refer to setting 2614 in Chapter 15.2.
Therefore set NO for this stage.
is set.
LSP2613.tif
15.5 I STUB stage
Fig. 44 Inverse stage settings, backup overcurrent
2640 Ip> Pickup:

The co-ordination of inverse time graded
protection can be applied effectively to ob-
tain reasonably fast and sensitive selective
protection. In this application, the inverse
stage is not used, so the setting here is in-
LSP2614.tif
finity, A.
2642 T Ip Time Dial:
Fig. 45 I STUB stage settings, backup overcurrent
As the setting of 2640 above is infinity (∞),
this setting is not relevant and left on the
default value of 0.50 s. 2630 Iph> STUB Pickup:
This stage may be used as a normal definite
2646 T Ip Additional Time Delay: time delay stage. In addition to this, it pro-
This stage may also be used as a further vides for blocking or release via binary in-
definite time delay stage by using this set- put. For certain applications (e.g.1 ½ CB) a
ting. As the setting of 2640 above is infinity STUB exists when the line isolator is open.
(∞), this setting is not relevant and left on By releasing this overcurrent stage via the
the default value of 0.00 s. mentioned binary inputs, a fast selective
2650 3I0p Pickup: fault clearance for faults on the STUB can
The co-ordination of inverse time graded be obtained. In this application, no such
protection can be applied effectively to ob- STUB protection is required, so this stage
tain reasonably fast and sensitive selective is disabled by applying an infinite pickup
protection. In this application, the inverse value with the setting A.
stage is not used, so the setting here is in- 2631 T Iph STUB Time delay:
finity, A. As the setting of 2630 above is infinity (∞),
2652 T 3I0p Time Dial: this setting is not relevant and left on the
As the setting of 2650 above is infinity (∞), default value of 0.30 s.
this setting is not relevant and left on the 2632 3I0> STUB Pickup:
default value of 0.50 s. This stage may be used as a normal definite
2656 T 3I0p Additional Time Delay: time delay stage. In addition to this, it pro-
This stage may also be used as a further vides for blocking or release via binary in-
definite time delay stage by using this set- put. For certain applications (e.g.1 ½ CB) a
ting. As the setting of 2650 above is infinity STUB exists when the line isolator is open.
(∞), this setting is not relevant and left on By releasing this overcurrent stage via the
the default value of 0.00 s. mentioned binary inputs, a fast selective
fault clearance for faults on the STUB can
2660 IEC Curve: be obtained. In this application, no such
During the device configuration (Chapter STUB protection is required, so this stage
4) the standard of the curves was selected is disabled by applying an infinite pickup
with parameter 0126 to be IEC. Here the value with the setting A.
choice is made from the various IEC
curves. As the stage is not applied in this 2633 T 3I0 STUB Time delay:
application the setting is not relevant and As the setting of 2632 above is infinity (∞),
left on the default value of Normal inverse. this setting is not relevant and left on the
default value of 2.00 s.

2634 Instantaneous trip via Teleprot./BI: 2915 Voltage Failure Supervision:

The I>> stage is applied for this purpose, In the event of energising the primary cir-
refer to setting 2614 in Chapter 15.2. cuit with the voltage transformer second-
Therefore set NO for this stage. ary circuit out of service, an alarm “168
Fail U absent” will be issued and the emer -
gency mode activated. This monitoring
task can be controlled with this parameter
2915. As no auxiliary contacts of the cir-
is set.
cuit-breaker are allocated, it is only con-
trolled with current supervision.
n 16. Measurement supervision – Setting
Group A
16.3 VT mcb
16.1 Balance / Summation
LSP2617.tif
Fig. 48 VT mcb settings, measurement supervision
2921 VT mcb operating time:

If an auxiliary contact of the mcb is utilised
(allocated in the matrix), the operating
LSP2615.tif
time of this contact must be entered here.

Note that such an input is not required in
Fig. 46 Balance/Summation settings, measurement practice as the relay detects all VT failures,
supervision including the operation of the mcb via
measurement. This set time will delay all
2901 Measurement Supervision: distance protection fault detection so it
Only in exceptional cases will the measure- should in general not be used. In this appli-
ment supervision not be activated. There- cation, it is also not required and therefore
fore this setting should always be ON. left on the default setting of 0 ms.
The advanced settings 2902A to 2909A can be
used to modify the parameters of the monitoring n 17. Earth-fault overcurrent – Setting Group A
functions. Generally, they can all be left on their 17.1 General
default values.
16.2 Measured voltage failure

LSP2618.tif
Fig. 49 General settings, earth fault overcurrent

LSP2616.tif
3101 Earth Fault overcurrent function is:

Fig. 47 Measured voltage fail settings, measurement For the clearance of high resistance earth
supervision
faults this function provides better sensitiv-
ity than the distance protection. As high re-
2910 Fuse Failure Monitor:
sistance earth faults are expected in this
Only in exceptional cases will the fuse fail-
application, this function is activated by
ure monitor not be activated. Therefore
setting it ON.
this setting should always be ON.
3102 Block E/F for Distance protection:
For the voltage failure detection, the default pa-
As the distance protection is more selective
rameters can be applied for the advanced settings.
(defined zone reach) than the earth-fault
protection and has superior phase selection
it is set to block the E/F protection with
every pickup.
3174 Block E/F for Distance Protection Pickup: 3109 Single pole trip with earth flt. prot.:
As fast single-pole tripping is only done The distance protection is set to cover all
with Zone 1 and Zone 1B with the distance arc faults on the line. High resistance faults
protection, the earth-fault protection is usually are due to mechanical defects (bro-
only blocked when the distance protection ken conductors or obstructions in the line)
picks up in Zone Z1/Z1B. so that an automatic reclosure is not sensi-
ble. Therefore, set earth fault only for
3103 Block E/F for 1pole Dead time:
three-pole tripping by application of NO.
During 1-pole dead times, load current can
flow via the zero sequence path. To prevent 3170 2nd harmonic for inrush restraint:
incorrect operation of the earth-fault pro- When energising the line, connected trans-
tection as a result of this it should be formers and load may cause an inrush cur-
blocked. Therefore set YES. rent with zero sequence component. This
rush current can be identified by its 2nd
3104A Stabilisation Slope with Iphase:
harmonic content. In this application in-
When large currents flow during faults
rush blocking is not required and not ap-
without ground, CT errors (saturation)
plied in the individual stages. The setting is
will cause current flow via the residual
of no consequence, so leave the default
path. The earth-fault protection, having a
value of 15 %.
very sensitive pickup threshold for high re-
sistance faults, could pick up due to this 3171 Max. Current, overriding inrush restraint:
CT error current. To prevent this, a stabi- If very large fault currents flow, CT errors
lising characteristic is provided to increase may also cause some 2nd harmonic. There-
the threshold when the phase currents are fore the rush blocking is disabled when
large. The characteristic is shown below in current is above this threshold. As stated
Figure 50. The default setting of 10 % is for parameter 3170 above, the inrush re-
suitable for most applications. straint is not applied in this example. This
setting is of no consequence, so leave the
default value of 7.50 A.
3172 Instantaneous mode after SwitchOnTo
Fault:
The earth-fault protection may be acti-
vated with a set time delay (parameter
3173) in the case of line energising (SOTF).
In this application, only the distance pro-
tection function is used for SOTF so that
this setting is of no consequence, so leave
the default value of with pickup and direc-
tion.
Fig. 50 Stabilisation of 3I0 pickup threshold
3173 Trip time delay after SOTF:
3105 3I0-Min threshold for Teleprot. schemes: As stated for parameter 3172 above, this
For directional comparison protection, in timer defines the delay of the SOTF trip by
particular for the weak infeed echo func- earth-fault protection. As it is not applied,
tion, a teleprotection send or echo block the default value of 0.00 s is left unchanged.
condition must be more sensitive than the
teleprotection trip condition. This thresh-
old determines the minimum earth current
for teleprotection send and is set to 80 % of
the most sensitive teleprotection trip stage.
Set
3I 0 _ TP = 0.8 ⋅ 3I 0 >>
3I 0 _ TP = 0.8 ⋅ 0.58
3I 0 _ TP = 0.46 A
Therefore apply the setting of 0.46 A.

17.2 3I0>>> 17.3 3I0>>
LSP2620.tif
LSP2619.eps
Fig. 52 3I0>> stage settings, earth fault overcurrent
Fig. 51 3I0>>> stage settings, earth fault overcurrent

3120 Operating mode:
In this application only three stages will be
3110 Operating mode: used, the 3I0>>> stage for fast (single time
A total of 4 stages are available, one of step) directional operation and the 3I0>>
which may be applied as inverse stage. In stage for time delayed directional opera-
this application only three stages will be tion and fast directional comparison as
used, the 3I0>>> stage for fast (single time well as the 3I0> stage for non-directional
step) directional operation and the 3I0>> backup operation. This stage must there-
stage for time delayed directional opera- fore be set to Forward.
tion and fast directional comparison as
3121 3I0>> Pickup:
well as the 3I0> stage for non-directional
This stage must operate for all internal
backup operation. This stage must there-
high resistance faults, use a 20 % margin.
fore be set to Forward.
3I 0 >> Pickup = 0.8 ⋅ I 1ph min_ R
3111 3I0>>> Pickup:
This stage must operate with the same sen- 3I 0 >> Pickup = 0.8 ⋅ 729 = 583 A
sitivity as the backup (emergency)
In secondary values therefore set 0.58 A.
overcurrent stage 3I0>> (refer to setting
2612). Therefore apply the setting 3.50 A 3122 T 3I0>> Time delay:
here. Note that this stage is only active This stage must operate with two time step
when distance protection is not picked up, delays. Therefore set 0.50 s here.
and it is directional (no reverse fault opera-
3123 Instantaneous trip via Teleprot./BI:
tion) whereas the backup O/C stage only
The stage 3I0>> will operate with
operates in the emergency mode when dis-
teleprotection, so the setting here is YES .
tance protection is not available.
3112 T 3I0>>> Time delay:
As stated above, only the distance protec-
This stage must operate with single time
tion operates with SOTF, so set NO here.
step delay. Therefore set 0.25 s here.
3125 Inrush Blocking:
As stated above, inrush blocking is not ap-
plied, so set NO here.
teleprotection, so the setting here is NO .
tion operates with SOTF, so set NO here.
3115 Inrush Blocking:
As stated above, inrush blocking is not ap-
plied, so set NO here.

17.4 3I0> 17.5 3I0 Inverse time
LSP2621.tif
Fig. 53 3I0> stage settings, earth fault overcurrent
LSP2622.tif
3130 Operating mode:
In this application only three stages will be Fig. 54 3I0 Inverse time stage settings, earth fault overcurrent
used, the 3I0>>> stage for fast (single time
step) directional operation and the 3I0>> 3140 Operating mode:
stage for time delayed directional opera- This stage is not required so it is set to
tion and fast directional comparison as Inactive.
well as the 3I0> stage for non-directional
backup operation. This stage must there- Because this stage is inactive, the settings 3141 to
fore be set to Non-Directional. 3151 are of no consequence and left on their de-
fault values.
3131 3I0> Pickup:
This stage must operate for all internal 17.6 Direction
high resistance faults, the same as 3I0>>,
but non-directional and with longer time
delay. In secondary values therefore set
0.58 A.
3132 T 3I0> Time delay:
This stage must operate with four time step
delays. Therefore set 1.00 s here.
LSP2623.tif
teleprotection, so the setting here is NO.
3134 Instantaneous trip after SwitchOnToFault: Fig. 55 Direction settings, earth fault overcurrent
tion operates with SOTF, so set NO here. 3160 Polarization:
Because both applied stages of the earth-
3135 Inrush Blocking: fault overcurrent protection are directional
As stated above, inrush blocking is not ap- (forward), the choice of polarising signal
plied, so set NO here. must be carefully considered. If both nega-
tive and zero sequence infeed are present at
the relay location polarisation with U0 +
IY or U2 provides excellent results. The
earth current from a star connected and
earthed transformer winding is only in-
cluded, when the 4th current input of the
relay is connected as such. In this applica-
tion, this current input measures the resid-
ual current of the protected line, (parame-
ter 220 in Chapter 7.1). Therefore only the
zero sequence or negative sequence voltage
are used as polarising signal with this set-
ting. The choice is automatic (the larger of
the two values is chosen individually dur-
ing each fault).

3162A ALPHA, lower angle for forward direction: 3169 Forward direction power threshold:
The default direction limits have been opti- This setting is only relevant for direction
mised for high resistance faults and are left decisions based on zero sequence power. In
unchanged at 338° here. this application it is of no consequence and
left on default value of 0.3 VA.
3163A BETA, upper angle for forward direction:
The default direction limits have been opti-
n 18. Teleprotection for earth fault
mised for high resistance faults and are left
overcurrent – Setting Group A
unchanged at 122° here.
3164 Min. zero seq. voltage 3U0 for polarizing:
The zero sequence voltage is one of the val-
ues for directional polarising. Under high
resistance fault conditions, this value may
LSP2624.tif
become very small. For the setting it is cal-
culated using the minimum single-phase
fault current under high resistance fault Fig. 56 Teleprotection for earth fault overcurrent
conditions and the smallest zero sequence settings
source impedance (this includes a safety
margin as these two conditions will not co- 3201 Teleprotection for Earth Fault O/C:
incide): In this application the teleprotection is re-
quired and applied as “Directional Com -
3I 0 min = I 1 ph min_ R ⋅ Z 0 source_ min parison Pickup”, refer to parameter 132 in
3I 0 min = 729 ⋅ 20 = 14.58 kV Chapter 4. This function is therefore acti-
vated by setting it ON.
As secondary value this is:
3202 Line Configuration:
100 V
3U 0 min_ sec = 3U 0 min ⋅ The line is a two terminal line.
380 kV
3203A Time for send signal prolongation:
100 V As the same type of communication with
3U 0 min_ sec = 14.58 kV ⋅ = 38
. V
380 kV the same channel delay time is used for dis-
tance teleprotection and earth-fault tele-
Therefore apply the setting 3.8 V. protection, the same setting consideration
3166 Min. neg. seq. polarizing voltage 3U2: as for parameter 2103A in Chapter 13 ap-
Although a similar calculation as done for plies here. Therefore in this example a set-
3164 would return a smaller value (50 %), ting of 0.05 s is applied.
this is not applied, as an automatic selec- 3209A Transient Block.: Duration external flt.:
tion of the larger of the two voltages was As the same type of communication with
set (parameter 3160). The setting applied the same channel delay time is used for dis-
here therefore is the same as that for the tance teleprotection and earth-fault tele-
zero sequence voltage. Therefore apply the protection, the same setting consideration
setting 3.8 V. as for parameter 2109A in Chapter 13 ap-
3167 Min. neg. seq. polarizing current 3I2: plies here. Therefore in this example a set-
Apply here the minimum negative se- ting of 0.04 s is applied.
quence current flowing for high resistance 3210A Transient Block.: Blk. T. after ext. flt.:
faults with a 20 % margin. As the same type of communication with
3I 2 min = 0.8 ⋅ I 1 ph min_ R the same channel delay time is used for dis-
tance teleprotection and earth-fault tele-
3I 2 min = 0.8 ⋅ 729 = 583.2 A protection, the same setting consideration
Therefore apply the setting 0.58 A. as for parameter 2110A in Chapter 13 ap-
plies here. Therefore in this example a set-
3168 Compensation angle PHI comp. for Sr: ting of 0.05 s is applied.
This setting is only relevant for direction
decisions based on zero sequence power. In
this application it is of no consequence and
left on default value of 255° .

n 19. Automatic reclosure – Setting Group A 3406 Evolving fault recognition:

19.1 General If during the single-pole dead time a fur-
ther fault is detected (evolving fault), the
AR function can respond to this in a de-
fined manner. The detection of evolving
fault in this application will be done by de-
tection of a further (new) trip command
initiation to the AR function. Therefore set
with Trip.
LSP2625.tif
3407 Evolving fault (during the dead time):
The response to the evolving fault during
Fig. 57 General settings, automatic reclosure the single-pole dead time is set here. In this
application it is set to starts 3pole AR-
3401 Auto-Reclose function: cycle.
In this application the automatic reclosure 3408 AR start-signal monitoring time:
function is required and applied with “1 If the AR start signal (protection trip) does
Cycle” and “with Trip and Action Time”, not reset after a reasonable time (breaker
refer to parameters 133 and 134 in Chap- operating time plus protection reset time),
ter 4. This function is therefore activated then a problem with either the circuit-
by setting it ON. breaker (breaker failure) or the protection
3402 CB ready interrogation at 1st trip: exists and the reclose cycle must not be
Before a reclosure is attempted the circuit- started. Here the maximum time for the
breaker status must be checked. This can initiate signal is set. If it takes longer, the
be done before the reclose cycle is started AR cycle is not started and a final trip con-
(prior/at time of initiation) or before the dition is set. Apply a setting of twice the
reclose command is issued. In this applica- breaker operating time plus protection re-
tion the breaker status is checked before set time, i.e. 0.20 s.
the close command is issued by the reclo- 3409 Circuit Breaker (CB) Supervision Time:
ser, so this setting must be NO. As the CB ready status will be checked
3403 Reclaim time after successful AR cycle: prior to issue of the close command, a time
If the reclose is successful, the recloser limit must be applied during which this
must return to the normal state which ex- ready status must be reached. If it takes
isted prior to the first fault. The time set longer, a final trip status is set and
here is started by each reclose command reclosure does not take place. This time
and must take the system conditions into limit is set here to be 3.00 s.
account (also the circuit-breaker recovery 3410 Send delay for remote close command:
time may be considered). Here a setting of The AR function can be applied to send a
3.00 s is applied. close command to the remote end via com-
3404 AR blocking duration after manual close: munication channels. This is not applied
If the manual close binary input is as- here, so the time is left on the default set-
signed, then the AR should be blocked for a ting of infinity, s.
set time after manual close to prevent AR 3411A Maximum dead time extension:
when switching onto a fault. In this appli- The AR function can be applied to wait for
cation the manual close binary input is not release by sync. check or CB status before
assigned, SOTF is recognised by current release of close command. Here, the maxi-
flow and AR is not initiated in this case. mum extension of the dead time in the
This setting is not relevant here as the course of waiting for release conditions is
manual close binary input is not assigned, set. In practice, a limitation to less than 1
the default setting of 1.00 s is left un- minute is practical. In this application 10 s
changed. will be used.

19.2 1st auto reclose cycle 19.3 3pTRIP / dead line charge / reduced dead
time
LSP2627.tif
LSP2626.tif
Fig. 59 3-pole Trip/DLC/RDT settings, automatic reclosure
Fig. 58 1st AR cycle settings, automatic reclosure

3430 3pole TRIP by AR:
If the AR function is initiated by single-
3450 Start of AR allowed in this cycle:
pole trip signals, it may in the course of a
As this is the only AR cycle that is applied,
single-pole AR cycle detect that the condi-
starting must be allowed in this cycle, so set
tions for single-pole reclosure are no lon-
this parameter to YES.
ger valid (e.g. because of further single-
3451 Action time: pole trip during the dead time or auxiliary
As indicated with parameter 134 in Chap- contact status from the breaker indicating
ter 4, the action time is used to differentiate multiple pole open, etc.). In such an event,
between faults cleared without delay by the the AR function may issue a three-pole trip
main protection and backup protection before proceeding with a three-pole AR cy-
operation for external faults. The action cle or setting the final trip condition. This
time must be set below the calculated coor- setting determines whether the AR func-
dination time step (0.25 s) and must be tion will issue such a three-pole trip. In this
longer than the slowest operation with application no external initiate signals are
teleprotection (60 ms). A time of 0.20 s is applied so this setting is set to NO, because
applied. the internal protection functions can man-
age their own three-pole coupling of the
3456 Dead time after 1pole trip:
trip signal when required.
The dead time must allow for the arc gases
to dissipate. During single-pole trip this 3431 Dead Line Check or Reduced Dead Time:
time is longer, because the arc is still sup- Special reclose programs can be applied to
plied by capacitive coupled current from prevent repetitive reclose onto fault and to
the healthy phases after the circuit-breaker minimise dead times. In this application
is open single pole. In practice a time of these programs are not used, so set With-
1.00 s has proven to provide good results. out.
3457 Dead time after 3pole trip: The settings 3438, 3440 and 3441 are of no conse-
The dead time must allow for the arc gases quence, because 3431 is set to “Without”.
to dissipate. During three-pole trip this Leave these settings on their default values.
time is short. In practice a time of 0.50 s
has proven to provide good results.
3458 Dead time after evolving fault:
As set in parameters 3407 and 3408 above,
a three-pole dead time will be started in the
case of evolving fault. Here the same time
as in parameter 3457 can be used because
this time is started with the three-pole trip
issued due to the fault evolving from single
to three phase. Therefore set 0.50 s.
3459 CB ready interrogation before reclosing:
As stated above, the CB status will be
checked before issue of close command.
Therefore set YES.
3460 Request for synchro-check after 3pole AR:
The sync. check condition must be checked
before issue of close command. Therefore
set YES.
19.4 Start autoreclose Where parallel lines can couple in larger

voltages onto the dead line, a higher setting
may be appropriate. In this case, no paral-
lel lines exist, so the setting of 10 % will be
used:
Udead = 0.1 ⋅ UN
LSP2628.tif
Udead = 0.1 ⋅ 110 = 11 V
Fig. 60 Settings for starting the AR, Automatic reclosure
The value for UN in this case (110 V) is
3420 AR with distance protection: based on the busbar voltage transformers
The distance protection will trip single pole (phase to phase), as this results in the larger
and three pole and start the autoreclos- (more conservative) setting. Apply the set-
ure, so set YES here. ting of 11 V.
3422 AR with weak infeed tripping: 3503 Voltage threshold live line/bus:
The weak infeed tripping will trip single If the measured line/bus voltage is above
pole and three pole and start the autoreclos- this set threshold, the line/bus is consid-
ure, so set YES here. ered to be switched on (live). In this appli-
cation the L3-L1 phase to phase voltage is
3423 AR with earth fault overcurrent prot.: used for synchronism check, so that the
The earth-fault overcurrent protection will voltage thresholds must be based on phase
trip three pole and not start the autoreclos- to phase voltage. The setting must be below
ure, so set NO here. (20 % safety clearance) the minimum an-
3425 AR with back-up overcurrent: ticipated operating voltage (in this example
The backup overcurrent protection will 85 % of the nominal voltage):
trip and start the autoreclosure, so set YES Ulive = 0.8 ⋅ 0.85 ⋅ UN
here. Note that due to the action time (pa-
400 kV
rameter 3451 in Chapter 19.2) only the ac- U live = 0.8 ⋅ 0.85 ⋅ ⋅100 = 716
. V
celerated trip with teleprotection will result 380 kV
in reclosure.
The value for UN in this case (400/380 ⋅110 V)
is based on the line voltage transformers
n 20. Synchronism and voltage check –
(phase to phase), as this results in the lower
Setting Group A
(more conservative) setting. Apply the set-
20.1. General ting of 71 V.
3504 Maximum permissable voltage:
If the measured line or bus voltage is above
this setting, it is considered to be a too
large operating voltage for release of a close
command. This setting must be above the
LSP2638.tif
highest expected operating voltage that is

still acceptable for release of the close com-
Fig. 61 General settings for synchronism and voltage mand. In general, a setting of 110 % of the
check normal operating voltage is recommended.
On long lines, the local line voltage fed
3501 Synchronism and Voltage Check function: from the remote end may rise to a higher
In this application the synchro check is value due to the Ferranti effect (normally
used, so this function must be selected ON. compensated by shunt reactors on the
3502 Voltage threshold dead line/bus: line). In this case, a higher setting may be
If the measured line/bus voltage is below required. In this example we will use a
this set threshold, the line/bus is consid- 110 % setting:
ered to be switched off (dead). In this ap- Umax = 1.10 ⋅ UN
plication the L3-L1 phase to phase voltage
is used for synchronism check, so that the Umax = 1.10 ⋅ 110 = 121 V
voltage thresholds must be based on phase The value for UN in this case (110 V) is
to phase voltage. In general, a setting of based on the busbar voltage transformers
10 % can be applied, when the voltage is (phase to phase), as this results in the larger
below 10 %, the line/bus is definitely dead. (more conservative) setting. Apply the set-
ting of 121 V.
3507 Maximum duration of synchronism check: 3510 Operating mode with AR:
If the synchronism check conditions are If the close release must be possible under
not obtained within this set time, the sync. asynchronous conditions (line and bus fre-
check is terminated without release or quency are not the same), then the circuit-
close. The person that issues the sync. breaker closing time must be considered
check request (operator close command for timing of the close command. Refer to
initiation) expects a response from the setting 239 in Chapter 7.3. In this example
switchgear within a reasonable time. This closing under asynchronous conditions
time should be set to the maximum time must be possible, so apply the setting: with
such operating personnel would accept to consideration of CB closing time.
wait for a response. Typically, a setting of
3511 Maximum voltage difference:
60.00 s is an acceptable delay.
In this setting the maximum voltage mag-
3508 Synchronous condition stability timer: nitude difference is set. If the magnitudes
When the set synchro check conditions are of the line and bus voltage differ by more
met, the release can be delayed by this set- than this setting, the sync. check function
ting to ensure that this condition is not will not release reclosure. As sync. check is
only a transient condition. In general, this done with phase to phase voltage in this
additional stability check is not required, case, the setting must be based on ph-ph
so that this setting can be set to 0.00 s. voltage. Use the difference between the
maximum and minimum operating volt-
3509 Synchronizable circuit breaker:
age to obtain the worst case result:
The integrated control functions may also
trigger the sync. check measurement. For UDiff max = (UN max - UN min)
this purpose the appropriate switchgear
 400 kV 
item can be selected in this setting. In this U Diff max = 121 − ⋅100 ⋅ 0.85 
application example, the integrated control  380 kV 
functions are not used, so that the setting
<none> is applied. UDiff max = 31.5 V
A setting of 31.5 V is under normal cir-
20.2. Settings for operation with auto-reclosure cumstances too large, as switching with
The following group of settings is relevant for such a large delta would cause a severe
close commands that originate from the transient in the system. Unless special cir-
auto-reclose function. This can be the internal AR, cumstances exist, such as very long lines
which is directly coupled with the internal sync. with Ferranti voltage rise or very weak in-
check, or an external AR device that couples the terconnections without voltage compensa-
trigger signal via binary input to the sync. check tion (e.g. tap changers), an upper limit of
function. approximately 20 % of the nominal voltage
should be applied:
UDiff max = 0.2 ⋅ UN
UDiff max = 0.2 ⋅ 110 = 22 V
Therefore, apply a setting of 22 V.
3512 Maximum frequency difference:
If the frequency difference between line
and bus voltage is less than this setting, the
sync. check conditions for async. switching
LSP3639.tif
can be used. The sync. conditions for

switching apply, if the frequency difference
Fig. 62 Sync. check settings for auto-reclose trigger
is less than 0.01 Hz. Switching with large
frequency difference will cause severe tran-
sients to the system. In common practice,
an upper limit of 0.10 Hz is suitable.

3513 Maximum angle difference: 3530 Operating mode with Man.Cl:

Under synchronous switching conditions If the close release must be possible under
(f diff < 0.01 Hz) the angle difference be- asynchronous conditions (line and bus fre-
tween the bus and line voltage is also quency are not the same), then the circuit-
checked. This angle under synchronous breaker closing time must be considered
conditions is stable and mainly due to the for timing of the close command. Refer to
transmission angle of the system. In this setting 239 in Chapter 7.3. In this example
example, a maximum angle of 40° will be closing under asynchronous conditions
applied. must be possible, so apply the setting: with
consideration of CB closing time.
3515A Live bus / live line and Sync before AR:
In this application, the auto-reclose func- 3531 Maximum voltage difference:
tion may initiate closing when bus and line The same consideration as for the AR clos-
voltage are live; therefore set YES. The ing in parameter 3511 applies. Therefore,
above sync. check conditions will then be apply a setting of 22 V.
monitored before closing is released.
3532 Maximum frequency difference:
3516 Live bus / dead line check before AR: The same consideration as for the AR clos-
In this application, the auto-reclose func- ing in parameter 3512 applies. Therefore,
tion may initiate closing to energise a dead apply a setting of 0.10 Hz.
line; therefore set YES.
3533 Maximum angle difference:
3517 Dead bus / live line check before AR: The same consideration as for the AR clos-
In this application, the auto-reclose func- ing in parameter 3513 applies. Therefore,
tion may not initiate closing to energise a apply a setting of 40° .
dead bus; therefore set NO.
3535A Live bus / live line and Sync before MC:
3518 Dead bus / dead line check before AR: In this application, the manual close may
In this application, the auto-reclose func- initiate closing when bus and line voltage
tion may not initiate closing to connect a are live; therefore set YES. The above sync.
dead bus to a dead line; therefore set NO. check conditions will then be monitored
before closing is released.
3519 Override of any check before AR:
The sync. check override is only used dur- 3536 Live bus / dead line check before Man.Cl.:
ing testing or commissioning. Therefore It is common practice to allow all closing
set NO. modes for manual close, so apply the set-
ting YES.
20.3 Settings for operation with manual close
and control 3537 Dead bus / live line check before Man.Cl.:
The following group of settings is relevant for Refer to setting 3536; therefore set YES.
close commands that originate from the manual 3538 Dead bus / dead line check before Man.Cl.:
close binary input or internal control function. Refer to setting 3536; therefore set YES.
3539 Override of any check before Man.Cl.:
The sync. check override is only used dur-
ing testing or commissioning. Therefore
set NO.
LSP2640.tif
Fig. 63 Sync. check settings for manual close and

control input trigger

n 21. Fault locator – Setting Group A n 23. General device settings
LSP2631.tif
LSP2629.tif
Fig. 66 General device settings

Fig. 64 Fault locator settings
0610 Fault Display on LED / LCD:
3802 Start fault locator with: The LED and LCD image/text can be up-
The fault locator can only provide mean- dated following pickup or latched with
ingful data for faults on the protected line trip. In this application the last pickup will
unless the downstream feeders are in a be indicated: Display Targets on every
pure radial configuration without interme- Pickup.
diate infeed. In this application, there is 0640 Start image Default Display:
infeed at the remote bus, so that the fault The LCD display during default conditions
locator data is only desired when the pro- (no fault detection or trip) can be selected
tection trips for internal faults. Therefore from a number of standard variants. Here,
apply the setting with TRIP. variant 1 is selected with the setting image 1.
3806 Load Compensation:
The result of the single ended fault location n 24. Time synchronization & time format
computation may be inaccurate due to the settings
influence of load angle and fault resistance.
This was described in conjunction with the
parameter 1307 in Chapter 11.1. For single
phase to ground faults and for phase to
phase faults without ground a load com-
pensated measurement can be applied to
achieve better results. This function does
not work under all conditions, and a fault
location output that closely resembles the
protection operation is desired in this ap-
plication, so the load compensation is
switched off by setting NO here.
LSP2632.tif
n 22. Oscillographic fault records

Fig. 67 Time synchronization and time format settings
Time synchronization settings can be ap-

plied here. Various sources for synchro-
nizing the internal clock exist as shown in
Figure 67.
LSP2630.tif
Fig. 65 Settings for the oscillographic fault recording
0402A Waveform Capture:

In this application a recording must be
saved during internal and external faults,
even if the relay does not trip. Therefore
apply the setting Save with Pickup to save
the recording every time the relay detects a
fault (picks up).
With settings 0403A to 0415, the length
and configuration of the oscillographic re-
cord can be set to match the user require-
ments.

n 25. Interface settings n 26. Password settings

25.1 Serial port on PC settings
LSP2636.tif
LSP2633.tif
Fig. 71 Settings for password access

Fig. 68 Settings for serial port on PC
Various password access levels can be applied as
The PC serial port configuration is shown here. shown in Figure 71.
No settings are required in this case.
25.2 VD address settings n 27. Language settings

LSP2634.tif
Fig. 69 Settings for VD address

LSP2637.tif
These addresses can be left on default values.

Fig. 72 Settings for language in the device
25.3 Operator interface settings
The language settings shown depend on the lan-
guages that are installed with the DIGSI device
driver on the PC.
LSP2635.tif
Fig. 70 Settings for operator interface
The settings here apply to the system interface, if

this is used.

n 28. Summary
SIPROTEC 7SA6 protection relays comprise the
functions required for overall protection of a line
feeder and can thus be used universally. The diver-
sity of parameterization options enables the relay
to be adapted easily and clearly to the respective
application using the DIGSI 4 operating program.
Many of the default settings can easily be accepted
and thus facilitate the work for parameterization
and setting. Already when ordering, economic
solutions for all voltage levels can be realized by
selection of the scope of functions.

Directional Phase
Overcurrent Protection
ANSI 67 with 7SA522 and
7SA6
n 1. Introduction
The implementation of a directional overcurrent
stage (ANSI 67) in the distance protection relays
LSP2310.eps
7SA522 and 7SA6 is possible via a simple coupling
of the distance protection directional stage (in this
case Zone 5) with one of the overcurrent stages (in
this example stage I>) in the relay.
The distance protection and the directional
overcurrent protection use the same measured
signals, phase current and phase voltage, but the
impedance measurement achieves both higher
sensitivity and higher selectivity.
This document illustrates how easily the ANSI 67
function can be implemented in the 7SA522 and
7SA6 by using the CFC logic.
n 2. General parameters for implementation of

directional overcurrent (ANSI 67)
The settings of the relay can be applied as required
by the application as usual. For the ANSI 67 func-
tion at least the functions 0112 Phase Distance,
0113 Earth Distance and 0126 Backup Overcur-
rent must be activated in the relay configuration
Fig. 1 SIPROTEC line protection 7SA522 and 7SA6
(Fig. 2):
Setting parameter Designation Example
of 67 Value
(secondary
1 A)
Pick-up threshold I67> pick-up 2.5 A
Time delay Time 67 0.5 s
Table 1 Parameters for directional overcurrent I67
Setting block: Relay configuration

The distance protection must be enabled (quadri-
lateral or MHO) for phase faults: Parameter 0112.
Backup overcurrent must be enabled (time
LSP2748.if
overcurrent IEC or ANSI): Parameter 0126
Fig. 2 Setting of functional scope

Matrix: Masking I/O (configuration matrix) 1201 Distance protection is ON

Assign the signal “3719 Distance forward” in the 1202 Phase current threshold for distance meas-
“Distance General” block to destination CFC. urement ≤ I67> pick-up (Table 1)
To calculate the minimum reach setting of the dis-
tance protection, the overcurrent characteristic is
shown in the impedance plane below in Fig. 6:
LSP2749.tif
Fig. 3 Masking of dist. signals in configuration matrix
Select that stage in the backup overcurrent that

must be directional (ANSI 67). For this stage as-
sign the corresponding blocking input signal to
source CFC. Therefore either:
7104 >Block O/C I>> or
7105 >Block O/C I> (used in this example) or
7106 >Block O/C Ip Fig. 6 Pick-up characteristic of overcurrent protection
in impedance plane
Multiple assignment is also possible.
From Fig. 6 it is apparent, that the largest circle ra-
dius must be determined to set the minimum im-
pedance reach so that a direction decision is avai-
lable when the overcurrent stage picks up. The
maximum circle radius is obtained when the
measured fault voltage together with fault current
LSP2750.tif
at pick-up threshold occurs. If the maximum op-

erating voltage (unfaulted) is used, a large safety
Fig. 4 Masking of overcurrent signals in configuration factor is already incorporated as the fault voltage
matrix will always be less than this.
n 3. Special setting for the distance protection U operating max
Circle _ radius =
The distance protection sensitivity must be greater 3 ⋅ (I 67 > pick − up)
than or equal to that of the required directional
overcurrent protection. This does not present any In this example the rated secondary VT voltage is
problems, as the directional overcurrent protec- 100 V. The maximum operating voltage is 110%
tion will be set less sensitive than the maximum of rated, so the following circle radius can be
load current while the distance protection is set calculated:
more sensitive than the smallest fault current. 100 ⋅11
.
Circle _ radius =
In the relevant setting group (e.g. Setting Group 3 ⋅ 2.5
A) the following distance protection settings must
be checked: Circle _ radius = 25.4 Ω
The setting 1243 R load, minimum load imped-
ance (ph-ph) must be greater than the calculated
circle radius. In practice, this is no problem be-
cause this setting is calculated with a similar equa-
tion using the maximum load current instead of
the overcurrent pick-up (I67> pick-up) which
must be greater than the maximum load current.
Therefore, this setting when applied is by nature
greater than the circle radius calculated here.
LSP2751.tif
Fig. 5 Distance protection function general setting

At least one of the set distance protection zones

with forward or non-directional reach must have
a reach greater than the circle radius. In this ex-
ample the Zone 5 has the largest reach, and it is set
non-directional. In this case the following must be
checked:
1341 Operating mode Zone 5 is
non-directional
LSP2752.tif
1342 R(Z5) resistance for ph-ph faults ≥
Circle radius (25.4 Ω) Fig. 8 Settings for backup overcurrent function (I> in this example)
1343 X+(Z5) reactance for forward direction ≥ n 5. Setting up the CFC logic
Circle radius (25.4 Ω) In the CFC logic, the absence of a forward detec-
If none of the applied zones has a sufficiently large tion by the distance protection (in this case
reach, a special zone must be selected for this pur- Zone 5) will be used to block the backup over-
pose. This zone must then be set as non-direc- current stage (in this example stage I<).
tional (for MHO set forward) with an X and R In the fast PLC, insert a negator and route the sig-
reach equal to the calculated circle radius (25. 4 Ω nal “Dis forward” to its input. Route the negator
in this example). The time delay of this zone can output to the relevant blocking input to the
be set to infinity (∞) to avoid tripping by the dis- backup overcurrent stage.
tance protection in this zone if required.
LSP2753.tif
LSP2754.tif
Zone settings of non-directional zone Fig. 9 Input signal allocation in the CFC logic
n 4. Setting the backup overcurrent stage n 6. Testing the ANSI 67 function

for ANSI 67 To test the directional overcurrent, a forward and
The backup overcurrent must be operated as a reverse fault condition were simulated with the
backup protection so that it is always active: Omicron® state sequencer. For both faults, the
fault current level was set to 10 % above pick-up
2601 Operating mode ON: always active
of the ANSI 67 function, 2.75 A. The angle be-
In the backup overcurrent proctection the follow- tween current and voltage for the forward fault
ing settings must be applied to the stage that is in- was 0° and for the reverse fault 180°.
tended for the ANSI 67 function. In this example
The fault recordings and trip log for these two
stage I> will be used:
cases are shown on the following page:
2620 Iph> pick-up value
set equal to required I67> pick-up (in this example
2.5 A)
2621 T Iph> time delay
set equal to the desired 67 time delay (in this ex-
ample 0.5 s)
2622 3I0> pick-up value
disabled by setting to infinity (∞)

The trip log (Fig. 10) shows the trip signal by the
backup O/C I> stage after 510 ms.
LSP2755.tif
Fig. 10 Trip log for forward fault

LSP2756.tif
Fig. 11 Trip log of reverse fault
n 7. Summary
The implementation of a directional overcurrent failure, a similar logic may be implemented
stage (ANSI 67) in the distance protection relays whereby the blocking signal is derived from the
7SA522 and 7SA6 is possible via a simple coupling Fuse Fail alarm (170).
of the distance protection directional release with
one of the overcurrent stages in the relay.
If one of the other backup overcurrent stages is re-
quired as an emergency protection in the event
that the distance protection is blocked due to fuse

Distance Protection with

Parallel Compensation
Environmental and cost consciousness are forcing

utilities to install more and more parallel lines.
The close arrangement of the transmission lines
leads to a higher fault rate and to influencing of
the measuring results. The considerable influence
exerted by parallel lines on the measuring results
(of up to 30 % in distance protection) and the re-
medial actions are considered in this application
example.
n 1. Explanation of the term parallel line

1.1 Parallel lines with common positive and
negative-sequence systems
The two parallel lines have the same infeed at both
line ends.
Fig. 1 Parallel lines with same infeed at both line ends
LSP2567.tif
In this arrangement where both systems are con- Fig. 3 Parallel lines over the Bosporus
nected with the same infeed, it is possible (for dis-
tance protection) to compensate the influence of
the parallel line. 1.2 Parallel lines with common positive-sequence
and independent zero-sequence system
Fig. 2 Parallel line with only one common infeed

Fig. 4 Parallel line with common infeed at a common tower
In this arrangement of parallel lines the effect can

only be compensated on one side. This arrangement of parallel lines does not influ-
ence the distance measurement.

1.3 Parallel lines with isolated positive and zero- phase-conductors models the three-phase system.
sequence systems Two lines in parallel are modelled by two parallel
single conductors with an earth return path, for
which the mutual reactance must be calculated. In
the case of lines with earth-wires, an additional
coupling results, which must be considered in the
calculations. The coupling impedance can be cal-
culated as follows:
π ⋅µ 0 ϑ Ω
ZM ' = ⋅ f + jµ 0 ⋅ f ⋅ l n
4 Dab  km 
Fig. 5 Parallel lines with separate infeed
µ 0 = 4 π ⋅10 −4  Ω ⋅
s 
 km 
This is the most unfavorable arrangement for dis-
tance protection. Compensation of the inductive ρ
ϑ = 658
coupling of the circuits is not possible. f
This arrangement causes a complicated fault volt- ϑ = Depth of penetration in ground
age and current distribution due to the inductive f = Frequency in Hz
coupling. ρ = Specific resistance in Ω / m
Dab = Spacing in meters between the two conductors
n 2. General
For a typical value of the specific earth resistance
of ρ = 100 Ω/m, a system frequency of 50 Hz,
a conductor spacing of 20 m and an earth-fault of
Ia = 1000 A, the following result is arrived at.
Z M ' = 0.05 + j 0.24 Ω / km
Then the induced voltage in the parallel conduc-
tor can be calculated with Ub = ZM ⋅ Ia, and 250 V
per km is obtained.
On a 100 km parallel line, this would give an in-
duced voltage in the conductor of 25 kV.
n 3. Calculation of the measuring error of the

distance protection caused by a parallel
line in the event of an earth fault
Fig. 6 Inductive coupling of parallel lines
When overhead lines follow parallel paths, a mu-

tual, inductive coupling of the current paths ex-
ists. In the case of transposed lines, this effect in
the positive and negative sequence system may be
neglected for all practical purposes (mutual reac-
tance less than 5 % of the self-impedance). This
implies that during load conditions, and for all
short-circuits without earth, the lines may be con-
sidered as independent. Fig. 7 Tower diagrams
During earth-faults, the phase currents do not add

up to zero, but rather a summation current corre-
sponding to the earth-current results. For this
summated current, a fictitious summation con-
ductor placed at the geometrical centre of the

Earth resistance The measured impedance ZB for the distance relay

100 Ω/ m Z2 is
Positive impedance (Ω / km) 0.032 + j 0.254 Z 0M
x⋅
Zero impedance (Ω / km) 0.139 + j 0.906 3⋅ Z L
Coupling impedance (Ω / km) 0.107 + j 0.488 Z 2 = ( 2 ⋅ l − x) ⋅ Z L +
Z
1+ E
R1 = 0.032 Ω / km ZL




0.254 Ω /km

X1 =
R0 = 0.139 Ω / km Measuring error
X0 = 0.906 Ω / km
R0M = 0.107 Ω / km
X0M = 0.488 Ω / km By placing the values in the equations we can cal-
culate the measuring errors for this double-circuit
ZE Z0 − Z1 line with single-end infeed.
= = 0.86
ZL 3 ⋅ Z1 The results are shown in the diagram below:
ZM Z0 − Z1
= = 0.65
ZL 3 ⋅ Z1
Fig. 8 Parallel line with an infeed without parallel line Fig. 9 Double-circuit line with single-end infeed
compensation
Phase current :
ILA = IA1+IA2+IA0 ILB = IB1+IB2+IB0
Earth current:
IEA= 3 IA0 IEB = 3 IB0
K0 = ( ZLo- ZL1 ) / 3 ZL1
K0M = Z0M / 3 ZL1
IC0 / IA0 = x / 2-x
For measured impedance ZA for the distance relay
Z1 is
Z 0M x
⋅
x x 3 ⋅ Z L 2l − x
Z1 = ⋅ZL + ⋅ZL
l l Z
1+ E
ZL







Measuring error
Fig. 10 Distance measuring error on a double-circuit line with

single-end infeed

The greatest measuring deviation (35 %) occurs in The measuring error of the relay on the faulty line
the event of a fault at the end of the line, because with two-end infeed is shown in the above dia-
the coupled length up to the fault position is at gram. It can be seen that the fault becomes nega-
maximum. tive in the case of faults in the first 50 % of the line
under the same infeed conditions. This is exactly
This example shows that the zone reach needs to
the reach where the earth current on the parallel
be reduced to 70 % to avoid overreaching in the
line flows in the opposite direction.
event of earth faults.
The following figures show that the parallel line
3.1 Result influence changes strongly with the switching state
n The fault is proportional to K0M = Z0M / 3 ZL1 of the parallel line. The reason is the different
n The fault increases with the ratio of the earth earth-current distribution.
current of the parallel line IEP to the earth-fault U ph− E
current of the relay Z ph− E =
I ph− E + kE ⋅ I E + kEM ⋅ I Ep
n The relay has an underreach when the earth-
fault current of the parallel line and the earth Z 0M

current of the relay are in phase (same direc- with kEM =
3 ⋅ Z 1L
tion)
n The relay has an overreach when the earth-fault
current of the parallel line and the earth current

of the relay have opposite phase (opposite direc-
tion).
Fig. 12 Fault at the line end
kEM
∆Ζ = ⋅ Z L Ⳏ24 % von ZL
1+ kE
Fault at the line end (Fig. 12):
Infeed sources for positive-sequence and zero-
sequence system at the line end
Fig. 13 Fault at the line end with open circuit-breaker
kEM
Fig. 11 Earth fault on a double-circuit line with two-end ∆Ζ = − ⋅ Z L Ⳏ24 % von ZL
infeed 1+ kE
Fault at the line end (Fig. 13):

One switch open, star-point earthing and relay at
opposite ends.

3 ⋅ kEM Z
∆Z = ⋅ Z L = 0M ⋅ Z L Ⳏ 40 % of ZL
1+ kE Z0
Fault at line end (Fig. 14):
Fig. 14 Infeed sources of the positive and zero-sequence systems at opposite

line ends
Z 0M
kEM ⋅
Z 0L
∆Z = -Z L ⋅ Ⳏ -10 % of ZL
1 + kE
Fault at line end (Fig. 15):
Fig. 15 Parallel line disconnected and earthed at both ends
n 4. Parallel line compensation

In order for the distance protection to be able to
operate with parallel line compensation, it is as-
sumed that it receives IEP of the parallel line as a
measuring variable.
UA
ZA =
I ph + kE ⋅ I E
 Z Z 
Z 1 L  I ph + EL ⋅ I E + 0 M ⋅ I Ep 
 Z 1L 3 ⋅ Z 1L 
=
I ph + kE ⋅ I E
As can bee seen from the equation the fault im-

pedance is measured correctly when we add the
Z0M
term ⋅ I EP in the denominator. With the
3 ⋅ Z1L
normal setting kE = ZE/ZL the denominator is then
reduced against the bracketed expression in the
counter and Z1L is produced as measured result.
The distance protection has another measuring
input to which the earth current of the parallel
line is connected. The addition is numeric. It
should be noted that the relay on the healthy line
sees the fault at too short a distance due to cou-
pling of the earth current of the parallel line. If
zone 1 of the line without a fault is set to 85 %, the
distance protection would lead to an overreach
through the fed parallel fault. The distance protec-
tion would still see faults on the parallel line at up
to 55 % line length in zone 1.
Fig. 16 Connection of the parallel line compensation

 ZE Z 
 = 0.86 / 0 M = 0.65 
 L
Z 3 ⋅ Z L 
ZL = line impedance
The so-called earth-current balance is used to pre-

vent this overfunction. It compares the earth cur-
rents of both line systems and blocks the parallel
line compensation when the earth current of the
parallel line exceeds the earth current of the own
line by a settable factor.
x
2−
IE1 2 ⋅l − x l
Fig. 17 Distance measurement with parallel line
= =
I E2 x x
compensation
l
At a setting of x/l of 85 %, the parallel line com-
pensation is effective for faults on the own line
and for a further 15 % into the parallel line. This
results in a factor of IE1 / IE2 = 1.35 as a standard
value for the earth-current balance.
Setting instructions for the parallel line compen-
sation
n The compensation is only possible where both
lines end in the same station.
n In the distance protection the compensation is
only used where no sufficient backup zone is
possible without compensation (When the dou-
ble-circuit line is followed by short lines).
Fig. 18 Effect of the parallel line compensation
Example:
S"SC = 10 000 MVA S"SC = 20 000 MVA

Z0 Z0
=1 =1
Z1 Z1
PN = 1000 MVA
ZVA1 = 16 Ω S"SC = 2500 MVA ZVC1 = 8 Ω
Fig. 19 Z0
Example of a =1
Z1
double-circuit line
station ZVB1 = 64 Ω

n 5. Calculation examples In the pickup zone the following lines should be in

The procedure for setting a normal single-circuit the protected zone in the worst switching state
line is explained in the available manuals. Special (single line follows parallel line). The following
applications are dealt with here. setting should be set for this:
. ⋅ ( Z A− B + 2 ⋅ Z B − C )
Z + AA = 11
5.1 Double-circuit line in earthed system
As a rule infeeds are available in the intermediate
The coupling in the zero-sequence system requires stations of the double-circuit line which have to
detailed consideration of the zone setting for earth be taken into account in the grading of the backup
faults. zones. This is demonstrated by the following ex-
5.2 General procedure ample (see Fig. 19):
It is recommended to first determine the grading Double-circuit line.
of the distance zones for phase faults without tak- Setting of the distance zones for phase-to-phase
ing into account the parallel line coupling. In the short-circuits
second step the zone reaches are then checked for
earth faults and a suitable earth-current compen- Given:
sation factor selected. 100 kV double-circuit line
The use of parallel line compensation must be Line data:
considered so that an adequate remote backup Configuration according to
protection can be ensured in the event of earth l1 and l2 = 150 km, l3 and l4 = 80 km
faults. Z1L'= 0.0185 + j 0.3559 Ω/km
Z0L'= 0.2539 + j 1.1108 Ω/km
5.3 Grading of the distance zones for Z0M'= 0.2354 + j 0.6759 Ω/km
phase-to-phase short-circuits Pnat. = 518 MW per line
The zones must be set according to the basic rules Current transformer: 2000/1 A
of grading plans. For the backup zones, the para- Voltage transformer: 400/0.1 kV
bolic course of the impedance dependent on the
Task:
fault location is important.
Calculation of the zone setting for relay D1.
When double-circuit lines are connected in series
there are also different reaches of the backup Solution:
zones dependent on the switching state and the Only X values are used in the short-circuit calcula-
infeed at the opposite end. tions, for the sake of simplicity:
Theoretically speaking, this results in relatively XL1 = XL2 = 0.3559 Ω/km · 150 km = 53.4 Ω
high effort for creating the grading plan of dou- XL3 = XL4 = 0.3559 Ω/km · 80 km = 28.5 Ω
ble-circuit lines. We generally apply a grading factor (GF) of 85 %.
The procedure is usually simpler in practice. The zone reaches are calculated as follows:
Half the impedance of the following parallel line X1 = 0.85 · 53.4 Ω
can be used for practical grading of the second For selective grading of the 2nd stage it is assumed
zone (double-circuit line follows single-circuit that the parallel line L2 is open but that always at
line). This gives: least half the short-circuit power is available from
the intermediate infeed in B. Grading takes place
Z 2A = GF 2 ⋅ (Z A⋅B + 0.5 ⋅ Z B ⋅C )
selectively to the end of the 1st zone of the distance
In the third zone, the grading should be per- relays of the following lines 3 and 4. That means
formed according to the backup protection strat- we can use about half the line impedance. This
egy. A selective grading for all switching states gives a simplified equivalent circuit. For a three-
leads to relatively short third stages which hardly pole fault in C we calculate the short-circuit cur-
get any longer than the corresponding second rents drawn in the figure.
stage. In the high and extra-high voltage system,
an attempt will be made for the third stage to
cover the following double-circuit line in normal
parallel line mode. In this case the following step
setting is derived:
. ⋅ ( Z A ⋅B + Z B ⋅C )
Z 3A = 11

5.4 Zone reach during earth faults

The earth-current compensation factor kE is de-
cisive in the Ph-E measuring systems. For single-
circuit lines this is set to the corresponding ZF/ZL
value. The protection then measures the same im-
pedance for Ph-Ph and earth faults.
On double-circuit lines the zero-sequence system
coupling produces a measuring error for earth
faults. The measurement can be corrected with the
parallel line compensation. This function is con-
tained optionally in the relays 7SA. Only the earth
current of the parallel line needs to be connected
to the relay and the coupling impedance set. The
earth-current balance may be left at the standard
value x/l = 85 %. The earth-current factor must be
Fig. 20 Protection setting in double-circuit lines: System adapted to the single-circuit line in this case.
data for the calculation example
5.5 Setting of the kE factor (operation without
parallel line compensation)
Under consideration of the intermediate infeed ef-
In the event that the parallel line compensation is
fect we get
not used, a kE factor must be found which ensures
 28.5  119 .  adequate protection for the possible operating
X 2 = 534
. + ⋅ 1 +  ⋅ 0.85 =
 2  .  
219 states of the double-circuit line (see Table 1).
64 Ω = 120 % X L 1
According to the above recommendation we get

for zone 3
X 3 = (534
. + 28.5) ⋅11
. = 901
. Ω = 169 % X L 1
and for the pickup zone:
X + A = (534
. + 2 ⋅ 28.5) ⋅11
. = 121 Ω = 226 % X L1
we adapt the reach of the zones in R direction to
the impedance of the natural power:
U N2 400 2
Z Nat. = = = 309 Ω
PNat. 518
We assume that a new line has to transmit double

the power temporarily and allow an additional
safety margin of 30 %. This gives the maximum R
reach of pickup as:
309
RA1 = 0.7 ⋅ = 108 Ω
2
We further select ϕA = 50° and
RA2 = 2 · RA1 = 208 Ω
An R/X ratio of 1 offers adequate compensation
for fault resistance for the distance zones.

x This results in:

a) This equation applies for ≤1
l
x
X' X E X 0M
GF1(1 + kXER ) + kXEM ⋅ 0M 1+ + ⋅ l
x
For >1 it applies that:
X '0L XL 3⋅ XL 2 − x
l 1+ kXEL U ⋅ sin ϕ SC x l
X Ph− E = Ph− E = ⋅ XL ⋅
 XE  l  XE 
I Ph +   ⋅ IE 1+  
 X'   XL  R  XL  R
b) kXEL =  EL 
 X '1 L  Line
x
 X '0 M 
c) kXEM =   1+
RE R0 M
+ ⋅ l
 3 ⋅ X '0L  Line RL 3 ⋅ RL 2 − x
U ⋅ cos ϕ SC x l
Adapting the setting to a particular operating state RPh− E = Ph− E = ⋅ RL ⋅
 RE  l  RE 
causes an overreach or underreach in the respec- I Ph +   ⋅ I E 1+  
tive other states. GF1 in % is the selected grading  RL  R  RL 
factor for the 1st zone (reach for phase-to-phase
faults). x/l in % then specifies how far the zone 1 Initially, only the X value measured is of interest
(Ph- E loop) reaches in the event of earth faults, for the reach.
referred to the line length.
XE X
Determining of the relay setting value kER is dem- With kXEL = and kXEM = 0 M this gives:
XL 3⋅ XL
onstrated by the example of double-circuit line
operation. For a fault at the distance x/l, the volt- x
age at the relay location is: l
1 + kXEL + kXEM ⋅
x x x Z0M x
U Ph⋅E = Z L ⋅ I Ph + Z E ⋅ I E + ⋅ I EP 2−
x l
l l l 3 X Ph− E = XL ⋅
l 1 + kXER
Whereby with single-end infeed:
x The Ph-E measuring system and the Ph-Ph mea-
suring system have the same impedance pickup
I Ph = I E and I EP = l ⋅ I E
x value (common setting value Z1).
2− Therefore: ZPh-E = ZPh-Ph = Z1 = GF1 · ZL, applies,
l
whereby GF1 is the grading factor of the first zone.
For the measurement on the Ph-E loop the result The following equation is finally arrived at for the
is earth-current compensation factor which must be
x set at the relay:
Z 0M
ZL + ZE + ⋅ l x
3 2− x
U Ph⋅E x l 1 + kXEL + kXEM ⋅ l
Z Ph− E = = ⋅ x
I Ph + kER ⋅ I E l 1 + kER 2−
kXER = l ⋅ x −1
kER is the complex earth-current compensation GF1 l
factor set at the relay.
We can vary the reach of the Ph-E measuring sys-
X and R are calculated separately for the numeri-
tems of a given zone reach for phase faults (GF1 in
cal relays 7SA. Simplified equations apply for this
% of ZL) by adjusting the kXER factor.
if phases and earth currents have the same phase
We can also solve the previous equation according
relation.
to x/l and then arrive at the reach for a given kXER
setting.
x [GF1 ⋅ (1 + kXER ) + 2(1 + kXEL )] − [..] − 8(1 + kXEL − kXEM ) ⋅ (1 + kXER ) ⋅ GF1
2
=
l 2 ⋅ (1+ kXEL − kXEM )
In the same way we derive the equations specified

in Table 1 for the cases “parallel line open” and
“parallel line open and earthed at both ends”.

Reach x/l at Ph-E short-circuits
x 1 + k XER x x (1 + k XER ) ⋅ GF1

= GF1 ⋅ = See equation on = a)
l 1+ k XEL l previous page l 1+ k X ' 0M
XEL − k XEM ⋅
X 0L
kXER-setting with: 85 % 71 % 108 %

x (75 %) (64 %) (98 %)
= 0.85
l 1 + k EL x
k ER = ⋅ − 1 = 0.71 ( 0.5 )
GF1 = 0.85 GF1 l
kXEL = 0.71
108 % 85 % 132 %
b) kXEM = 0.64
c) X'0M = 0.72 Ω/km
x /l
X'0L = 1.11 Ω/km 1 + k XEL + k XEM ⋅
k XER = 2 − x /l − 1
X'1L = 0.356 Ω/km GF1
kXER = 1.18
65 % 56 % 85 %
X ' 0M
1 + k XEL + k XEM ⋅
X ' 0L x
k XER = ⋅ − 1 = 0.31
GF1 l
Distance measurement in the event of earth faults: Reach (in X direction) dependent on the relay setting
X 
k XER =  E  and the switching state
 X L  Relay
The choice of the setting of kXER requires a com- the line always has the opposite direction, i.e. the
promise which takes all three cases of operation zone reach is increased. Reliable fast disconnec-
into account (Table 1)1). At a grading factor of tion can always be ensured by intertripping. How-
GF1 = 85 %, adaptation to the single-circuit line ever, in reducing the kXER factor it must be taken
usually offers an acceptable solution. The two-end into account that the reach of the backup zones is
disconnection of a line with earthing at both ends also reduced accordingly in the event of earth
only occurs in maintenance work, so the brief faults. Zone reach reduction (e.g. GF1 = 0.8)
overreach of 8 % is only rarely effective because should therefore also be considered instead of
overreaching is usually reduced by intermediate only reduction of the kXER factor.
infeeds.
5.6 Setting the overreach zone
In operation with single-pole auto-reclosure, the
overreach would only lead to excessive auto- Zone Z1B should be set to 120 – 130 % Z L. This
reclosure and not to final disconnection, provided reach would also apply for earth faults in the case
that a transient short-circuit is concerned (about of operation with parallel line compensation.
90 % of faults).
Alternatively, the reach can be reduced slightly for
earth faults by setting a lower kXER factor. If it were
reduced from kXER = 0.71 to kXER = 0.5, overreach
would just about be avoided in this example. The
reach with both lines in service would then only
be 64 %, taking into consideration that the paral-
lel line coupling only takes full effect in the worst
case of single-end infeed. In the normal case of 1) The numeric values in Table 1 were calculated with the
line layers of the previous example. For the sake of sim-
two-end infeed the earth current on the parallel plicity the complex factors kEL = 0.71 - j0.18 and
line is much lower in the event of faults close to kEM = 0.64 - j0.18 were only taken into account with
the middle of the line, and the zone reach corre- their real components, which correspond to the values
kXEL = XE/XL and kXEm = XM/(3 · XL) in the first approxi-
sponds almost to the single-circuit line. In addi- mation. This gives sufficient accuracy for the extra
tion, the parallel line coupling at the other end of high-voltage system.

Adapting the setting to an operating state causes x

an overreach or underreach in the respective other l
1 + kXEL + kXEM ⋅
states. GF1 in % is the selected grading factor for x
2−
the 1st zone (reach for Ph-Ph faults). x/l in % then x l
X Ph− E = XL ⋅
specifies how far zone 1 (Ph-E loop) reaches in the l 1 + kXER
event of earth faults, referred to the line length.
The Ph-E measuring system and the Ph-Ph mea-
Determining of the relay setting value kER is dem- suring system have the same impedance pickup
onstrated by the example of double-circuit line value (common setting value Z1).
operation: Therefore: ZPh-E = ZPh-Ph = Z1 = GF1 · ZL, applies,
The voltage at the relay location for a fault at the whereby GF1 is the grading factor of the first zone:
distance x/l is: The following equation finally results for the
x x X Z 0M earth-current compensation factor which must be
U Ph− E = Z L ⋅ I Ph + Z E ⋅ I E + ⋅ I EP set in the relay:
l l l 3
with single-end infeed: x
1 + kXEL + kXEM ⋅ l
x x
l 2−
I Ph = I E and I EP = ⋅ IE kXER = l ⋅ x −1
x
2− GF1 l
l
Without parallel line compensation the 120 %
for the measurement of the Ph-E loop the follow- reach must be dimensioned for the case of parallel
ing is derived: line operation under consideration of the previ-
x ously defined kXER factor.
Z 0M
ZL + ZE + ⋅ l x
3 2− x l
U Ph− E x 1 + kXEL + kXEM ⋅
= = ⋅ l x
Z Ph− E 2−
I Ph + kER ⋅ I E l 1 + kER l ⋅x
GF =
1+ kXER l
whereby kER is the complex earth current compen-
sation factor set on the relay. X and R are calcu- For a fault at the end of the line (x/l = 100 %) and
lated separately in the numerical relays 7SA. a safety margin of 20 %, the following equation
Simplified equations apply for this when phases for the overreach zone is produced:
and earth currents have the same phase angle.
120%
This results in X1B = GF100% ⋅ X L ⋅
100
x 1 + kXEL + kXEM
X E X 0M = ⋅ X L ⋅1.2
1+ + ⋅ l 1 + kXER
XL 3⋅ XL 2 − x
U Ph− E ⋅ sinϕ SC x l The selected kXER = 0.71 produces
X Ph− E = = ⋅ XL ⋅
 XE  l  XE  X1B = 165 % XL.
I Ph +   ⋅ IE 1+   Without parallel line compensation, the
 XL  R  XL  R
overreach zone must therefore be set very
high, so that a safety margin of 20 % is en-
sured in double-circuit line operation.
x
RE R0 M 5.7 Reach of the backup zones for earth
1+ + ⋅ l
RL 3 ⋅ RL 2 − x faults
U ⋅ cosϕ SC x l We observe the behavior of the distance
RPh− E = Ph− E = ⋅ RL ⋅
 RE  l  RE  measurement with and without parallel line
I Ph +   ⋅ I E 1+  
 RL  R  RL  R compensation.
Only the measured value is initially for the reach:

XE X
With kXEL = and kXEM = 0 M this gives:
XL 3⋅ XL

5.8 Distance measurement without parallel line

compensation
For the simple case that the parallel line is fol-
lowed by a single-circuit line (Fig. 21, line 4 dis-
connected), the measured impedance can be
determined as follows:
Voltage at the relay location:
Z 0M 1− 2
U Ph− E = Z L 1 ⋅ I Ph1 + Z E1 ⋅ I E 1 + ⋅ IE2
3
x x
+ Z L2 ⋅ I Ph3 + Z E2 ⋅ I E3
l2 l2
With IPh1 = IE1 = IE2 = ISC and IPh3 = IE3 = 2 · ISC

we get for the relay reactance:
U Ph− E ⋅ sinϕ SC
X Ph− E = =
I Ph1 + kXER ⋅ I E1
1 + kXEL 1 + kXEM1 − 2 x 1 + kXEL3
⋅ X L1 + 2 ⋅ ⋅ ⋅ X L2
1 + kXER l 2 1 + kXER
Fig. 21 Distance measurement on double-circuit lines: Fault on a following line
If we set x/l = 0 we arrive at the measured

reactance in the event of a fault at the opposite
station. For the calculated example this gives the
value
1 + 0.71 + 0.64
X Ph− E = = 1.37 ⋅ X L 1
0.71
This shows that in the case at hand the backup
zones only reach beyond the next station when
they are set greater than 137 % ZL1. This does not
apply for the 2nd zone at the selected setting.
At the grading (120 %) selected on the basis of the
phase-to-phase short-circuits, the 2nd zone would
only reach up to 91 % ZL1 in the parallel line state
in the event of an earth fault.
This problem is particularly pronounced when the
downstream line, according to which the second
zone has to be graded, is substantially shorter than
the protected line, and when only a short interme-
diate infeed is present.

Z 0M 1− 2 x x x Z 0M3 − 4
U Ph− E = Z L1 ⋅ I Ph1 + Z E 1 ⋅ I E1 + ⋅ I E 2 + Z L2 ⋅ I Ph3 + Z E2 ⋅ I E3 + ⋅ I E4
3 l2 l2 l2 3
With IPh1 = IE1 = IE2 = ISC and
 x x
IPh3 = IE3 =  2 −  ⋅ ISC and IE4 = ⋅ ISC
 l 2 l 2
we get:
2
x x  x
 2 −  ⋅ (1 + kXEL2 ) +   ⋅ kXEM3 − 4
1 + kXEL1 + kXEM1 − 2 l  l2   l2 
X Ph− E = ⋅ X L1 + 2 ⋅ X L2
1 + kXER 1 + kXER
Resolution according to x/l2 produces again the

equation for the reaches of the zones:
x 2(1 + kXEL2 ) − 4(1 + kXEL2 ) − 4(1 + kXEL2 − kXEM3 − 4 ) ⋅ ∆

2
=
l2 2 ⋅ (1+ kXEL2 − kXEM3 − 4 )
with
X L1  X 
∆= ⋅ (1 + kXER ) ⋅ Zone − (1 + kXEL1 + kXEM1 − 2 )
X L2  X L1 
For zone 3 (169 % XL1) we get x/l2 = 33 %, i.e. only

slightly more than for the single-circuit line. For the
pickup zone (226 % XL1) the expression under the
root is negative because the zone only reaches to
just past the next substation.
The limit (root = 0) is at 223 % XL1.
5.9 Distance measurement with parallel line

compensation
With parallel line compensation, faults on the
own line are measured in the correct distance.
For faults beyond the next substation the zones
are extended by the factor:
1 + kXER + kXEMR
k=
1 + kXER
according to the compensation factors set at the

relay.
The term 1 + kXER must be replaced in the equa-
tions on pages 12 and 13 by 1 + kXER + kXEMR.
This gives us a reach of up to 71 % ZL2 for the 2nd
zone, i.e. the zone reaches up to just before the
end of the first zone of the following line which is
set to 85 % ZL2.

Taking account of the intermediate infeed in sta-

tion B, there will be a further reduction of the 2nd
stage so that the safety margin is increased.
The 3rd zone (169 %ZL1) just reaches with parallel
line compensation and the pickup zone (226 %
ZL1) comfortably reaches over the next station but
one ( C ) (A fault in C would correspond to 162 %
ZL1). For the final definition of the setting, the in-
termediate infeed again has to be taken into ac-
count.
n 6. Summary
The zone setting can be estimated for the double-
circuit lines based on the arithmetic procedures
and the derived equations shown here. In practice
the intermediate infeeds must be taken into ac-
count, so that the second zone can be graded
safely past the next station (whilst retaining the se-
lectivity and reliably detecting busbar faults).
If the line lengths do not differ, an acceptable
compromise for the relay setting can usually be
found without parallel line compensation. For
short following lines, parallel compensation must
however be taken into account.
Computer programs are nowadays available for
the relatively complex testing of the backup zones
and the pickup.

Protection of Long Lines

with SIPROTEC 7SD5
n 1. Introduction
The protection of long transmission lines was
previously the domain of distance protection.
Modern available information transmission tech-
nology – with the ability to reliably exchange com -
parison signals over substantial distances – makes
differential protection interesting for use on long
transmission lines. High sensitivity and strict se-
LSP2173f.eps
lectivity are further aspects that speak in favor of

differential protection. The SIPROTEC 7SD5 relay
provides, in addition to differential protection,
comprehensive backup protection and additional
LSP2314_afp.eps
functions for the complete protection of transmis-
sion lines.
This application example describes largely differ-
ential protection of two-end lines. In addition to
this use, modern SIPROTEC differential protec- Fig. 1 SIPROTEC 7SD5 line differential protection relay
tion relays can meet the following requirements:
n Protection of multi-branch configurations
n Transformer in the protected zone
n Matching to various transmission media such as
fiber optics or digital communication networks
For protection of a two-end line, we recommend
to activate the following functions:
ANSI 87 L Differential protection
ANSI 67 N Directional overcurrent protection
ANSI 79 Auto-reclosure
ANSI 50 BF Breaker failure protection
ANSI 59/27 Undervoltage and overvoltage
protection
ANSI 25 Synchro-check and voltage check
LSP2747.tif
Fig. 2 Settings for functional scope

*
*
* Option
Fig. 3 Function scope of the SIPROTEC 7SD5
2.1 Differential protection

Differential protection is based on current com-
parison. It makes use of the fact, that e.g. a line
section L (Fig. 5) carries always the same current I
at its two ends in fault-free operation. This cur-
rent flows into one side of the considered zone
and leaves it again on the other side. A difference
in current is a clear indication of a fault within
this line section. If the actual current transforma-
Fig. 4 Differential protection for a line with two ends (single-phase system) tion ratios are the same, the secondary windings of
the current transformers CT1 and CT2 at the line
ends can be connected to form a closed electric
circuit with a secondary current I; a measuring
element M which is connected to the electrical ba-
lance point remains at zero current in fault-free
operation.
When a fault occurs in the zone limited by the
transformers, a current i1 + i2 which is propor-
tional to the fault currents I1 + I2 flowing in from
both sides is fed to the measuring element. As a
result, the simple circuit shown in Fig. 5 ensures a
reliable tripping of the protection if the fault cur-
rent flowing into the protected zone during a fault
Fig. 5 Basic principle of the differential protection for a line with two ends
is high enough for the measuring element M to
pick up.

2.2 Charging current compensation

Charging current compensation is an additional
function for differential protection. It allows an
improvement in the sensitivity, achieved by the
charging current (caused by the capacitances in
the overhead line or cable and flowing through the
distributed capacitance in steady state) being
compensated. As a result of the capacitances of the
phase conductors (to earth and mutually), charg-
ing currents flow – even in fault-free conditions –
and cause a difference in the currents at the ends Fig. 6 Charging current compensation for a line with two ends
of the protected zone. Particularly on cables and (single-phase system)
long lines the capacitive charging currents can
reach considerable levels. If the feeder-side trans- The directional polarization using the transformer
former voltages are connected to the relays, the starpoint current is independent of voltage trans-
influence of the capacitive charging currents can formers and therefore also functions reliably dur-
be largely arithmetically compensated. It is pos- ing a fault in the voltage transformer secondary
sible here to activate charging current compensa- circuit. It is, however, a requirement that not all,
tion, which determines the actual charging cur- but at least a substantial amount of the earth-fault
rent. Where there are two line ends, each relay current flows via the transformer, the starpoint
attends to half the charging current compensa- current of which is measured.
tion; where there are M relays, each covers an Mth
For the determination of direction, a minimum
fraction. Fig. 6 shows a single-phase system, for
current 3I0 and a minimum displacement voltage,
the sake of simplicity.
which can be set as 3U0>, are required. If the dis-
In fault-free operation, steady-state charging cur- placement voltage is too small, the direction can
rents can be considered as practically constant, as only be determined if it is polarized with the
they are determined only by the voltage and the transformer starpoint current and this exceeds a
line capacitances. Without charging current com- minimum value corresponding to the setting IY>.
pensation, they must therefore be taken into ac- The direction determination with 3U0 is inhibited
count when setting the sensitivity of the differ- if a “trip of the voltage transformer mcb” is repor-
ential protection. With charging current compen- ted via binary input.
sation, there is no need to consider them at this
point. With charging current compensation, the
steady-state magnetization currents across shunt
reactances (quadrature-axis reactances) are also
taken into account.
2.3 Directional overcurrent protection (ANSI 67 N)

The zero-sequence current is used as measured
quantity. According to its definition equation it is
obtained from the sum of three phase currents,
i.e. 3I0 = IL1 + IL2 + IL3.
The direction determination is carried out with
the measured current IE (= –3 I0), which is com-
pared to a reference voltage UP.
The voltage required for direction determination
UP may be derived of the starpoint current IY of an
earthed transformer (source transformer), provid-
ed that the transformer is available. Moreover,
both the zero-sequence voltage 3U0 and the star- Fig. 7 Directional characteristic of the earth-fault
point current IY of a transformer can be used for protection
measurement. The reference magnitude UP then is
the sum of the zero-sequence voltage 3U0 and
a value which is proportional to reference current
IY. This value is about 20 V for rated current
(Fig. 7).

2.4 Auto-reclosure function (ANSI 79)

The 85 % of the arc faults on overhead lines are The 7SD5 can also operate in conjunction with an
extinguished automatically after being tripped by external auto-reclosure device. In this case, the
the protection. This means that the line can be signal exchange between 7SD5 and the external
reclosed. Auto-reclosure is only permitted on reclosure device must be effected via binary inputs
overhead lines because the option of automatic and outputs. It is also possible to initiate the inte-
extinguishing of an arc fault only exists there. It grated auto-reclose function by an external pro-
should not be used in any other case. If the protec- tection device (e.g. a backup protection). The use
ted object consists of a combination of overhead of two 7SD5 with auto-reclosure function or the
lines and other equipment (e.g. overhead line di- use of one 7SD5 with an auto-reclosure function
rectly connected to a transformer or a combinati- and a second protection with its own auto-
on of overhead line/cable), it must be ensured that reclosure function is also possible.
reclosure can only be performed in the event of a Reclosure is performed by an auto-reclosure
fault on the overhead line. If the circuit-breaker function (AR). An example of the normal time
poles can be operated individually, a single-phase sequence of a double reclosure is shown in Fig. 8.
auto-reclosure is usually initiated for single-phase The integrated auto-reclosure cycle allows up to
faults and a three-pole auto-reclosure for multiple 8 reclose attempts. The first four reclose cycles
phase faults in the system with earthed system may operate with different parameters (action and
starpoint. If the earth fault still exists after auto- dead times, single/three-pole). The parameters of
reclosure (arc has not disappeared, there is a me- the fourth cycle also apply for the fifth cycle and
tallic fault), then the protection functions will re- onwards.
trip the circuit-breaker. In some systems several
reclosing attempts are performed.
In a model with single-pole tripping, the 7SD5
allows phase-selective, single-pole tripping. A
single and three-pole, single and multiple-shot
auto-reclosure function is integrated, depending
on the version.
Fig. 8 Timing diagram of a double-shot reclosure with action time (2nd reclosure successful)

2.5 Breaker failure protection (ANSI 50BF) In the latter case, the relay determines the time for
The breaker failure protection provides rapid ba- issuing the close command such that the voltages
ckup fault clearance, in the event that the circuit- are identical the instant the circuit-breaker poles
breaker fails to respond to a trip command from a make contact.
protection function of the local circuit-breaker. The synchronism and voltage check function uses
Whenever, e.g., a protection relay of a feeder issu- the feeder voltage – designated with ULine – and
es a trip command to the circuit-breaker, this is the busbar voltage – designated with UBus – for
repeated to the breaker failure protection. A timer comparison purposes. The latter may be any
in the breaker failure protection is started. The ti- phase-to-earth or phase-to-phase voltage.
mer runs as long as a trip command is present and
current continues to flow through the breaker po-
les.
2.6 Undervoltage and overvoltage protection

(ANSI 27/59)
Voltage protection has the function of protecting
electrical equipment against undervoltage and
overvoltage. Both operational states are unfavor-
able as overvoltage may cause, for example, insu-
lation problems or undervoltage may cause
stability problems.
The overvoltage protection in the 7SD5 detects the
phase voltages UL1-E, UL2-E and UL3-E, the phase-
to-phase voltages UL1-L2, UL2-L3 and UL3-L1, as well
as the displacement voltage 3U0. Instead of the
displacement voltage any other voltage that is Fig. 9 Synchronism check on closing
connected to the fourth voltage input U4 of the
relay can be detected. Furthermore, the relay cal-
culates the positive-sequence voltage and the ne- If a power transformer is located between the
gative-sequence voltage, so that the symmetrical feeder voltage transformers and the busbar voltage
components are also monitored. Here, compoun- transformers (Fig. 10), its vector group can be
ding is also possible which calculates the voltage at compensated for by the 7SD5 relay, so that no ex-
the remote line end. ternal matching transformers are necessary.
The undervoltage protection can also use the The synchronism check function in the 7SD5 usu-
phase voltages UL1-E, UL2-E and UL3-E, the phase- ally operates in conjunction with the integrated
to-phase voltages UL1-L2, UL2-L3 and UL3-L1, as automatic reclose, manual close, and the control
well as the positive-sequence system voltage. functions of the relay. It is also possible to employ
an external automatic reclosing system. In such a
2.7 Synchro-check and voltage check (ANSI 25) case, signal exchange between the devices is ac-
complished via binary inputs and outputs.
The synchro-check and voltage check functions
ensure, when switching a line onto a busbar, that When closing via the integrated control function,
the stability of the system is not endangered. The the configured interlocking conditions may have
voltage of the feeder to be energized is compared to be verified before checking the conditions for
to that of the busbar to check conformances in synchronism. After the synchronism check grants
terms of magnitude, phase angle and frequency the release, the interlocking conditions are not
within certain tolerances. Optionally, deenergiza- checked a second time.
tion of the feeder can be checked before it is con-
nected to an energized busbar (or vice versa).
The synchro-check can either be conducted only
for auto-reclosure, only for manual closure (this
includes also closing via control command) or for
both cases. Different close permission (release)
criteria can also be programmed for automatic
and manual closure. Synchronism check is also
possible without external matching transformers
if a power transformer is located between the
measuring points. Closing is released for synchro-
nous or asynchronous system conditions.

It is very important to note, however, that the

tripping times of the different protection relays
depend on the transmission quality and are pro-
longed in case of a reduced transmission quality
and/or an increased transmission time. Fig. 11
shows some examples for communication con-
nections. In case of a direct connection the dista-
nce depends on the type of the optical fiber.
Table 1 lists the options available. The modules in
the relays are replaceable. If an external communi-
cation converter is used, the relay must be equip-
ped with an FO5 module in order to achieve
correct operation. The relay and the external com-
munication converter are linked via optical fibers.
The converter itself allows connections to com-
Fig. 10 Synchronism check across a transformer munication networks, two-wire copper lines or
ISDN (Fig. 12).
Communications media To span larger distances with fiber-optic cables, it
The exchange of signals between the line differen- is presently recommended to use external repeat-
tial protection relays becomes particularly compli- ers. Optical modules for distances of up to 100 km
cated due to the geographic coverage of the relays are being developed and will be available in 2005.
over medium and long distances. At the mini- A further option is the connection via communi-
mum, two substations find themselves using re- cation network (no limitation of distance).
lays of the same system that exchange data over a
relay-to-relay (R2R) communication channel.
Fig. 11 Optional communication media
The communication is enabled via pilot wire or

direct optical fiber connections or via communi-
cation networks. Which kind of FO media is used,
depends on the distance and on the communicati-
on media available (Table 1). For shorter distances
a direct connection via optical fibers having a
transmission rate of 512 kBit/s is possible. A trans-
mission via modem and communication networks
can also be realized.

Module in the Connector FO type Optical Perm. path Distance,

relay type wavelength attenuation typical
FO5 ST Multi-mode 820 nm 8 dB 1.5 km
62.5/125 µm (0.95 miles)
FO6 ST Multi-mode 820 nm 16 dB 3.5 km
62.5/125 µm (2.2 miles)
FO7 ST Mono-mode 1300 nm 7 dB 10 km
9/125 µm (6.25 miles)
FO8 FC Mono-mode 1300 nm 18 dB 35 km
9/125 µm (22 miles)
1)
FO17 LC Mono-mode 1300 nm 13 dB 24 km
9/125 µm (14.9 miles)
1)
FO18 LC Mono-mode 1300 nm 29 dB 60 km
9/125 µm (37.5 miles)
FO191) LC Mono-mode 1550 nm 29 dB 100 km
9/125 µm (62.5 miles)
Table 1 Communication via direct FO connection
1) For direct connection over short distances, a

suitable optical attenuator should be used to avoid
malfunctions and damage to the relay.

n 3. Summary
Optimal protection of transmission lines with
SIPROTEC 7SD5 relays means high selectivity in
fault clearing; any available parallel duplicate line
remains reliably in operation. Very short release
times ensure the stability of the transmission
system in the event of a fault, and consequently
contribute significantly to maximizing the level
of supply security.
The SIPROTEC 7SD5 relay provides comprehen-
sive main and backup protection of transmission
lines in a single relay. Thanks to its flexible com-
munication capabilities, the SIPROTEC 7SD5
can be easily matched to the existing communi-
cation infrastructure.
Fig.12 Examples for communication links

Protection of a Three-
Winding Transformer
Three-winding transformer
110 kV/25 kV/10 kV
Yyn0d5
25 kV-side: Solidly earthed
Protection functions
ANSI 87 T - Differential protection
ANSI 87 N - Earth-fault differential protection
ANSI 50/51 - Definite-time overcurrent-time
LSP2309.eps
protection as backup
ANSI 49 - Thermal overload protection
ANSI 46 - Load unbalance protection
(negative-sequence protection)
ANSI 24 - Overexcitation protection
n 1. Introduction Fig. 1 SIPROTEC Transformer protection relay

Transformers are valuable equipment which make
a major contribution to the supply security of a differential protection relay 7UT613 can only be
power system. Optimum design of the transfor- used as backup protection against external faults
mer protection ensures that any faults that may in the connected power system. The backup pro-
occur are cleared quickly and possible consequen- tection for the transformer itself must be provided
tial damage is minimized. as a separate overcurrent relay (e.g. 7SJ602). The
In addition to design notes, a complete setting ex- Buchholz protection as fast short-circuit protec-
ample with SIPROTEC protection relays for a tion is delivered with the transformer.
three-winding transformer in the transmission
Designations in accordance with ANSI (American
system is described.
National Standard) are used for the individual
functions. The differential protection therefore
has the ANSI No. 87 for example.
The range of high-voltage transformers comprises The 7UT613 differential protection relay is pro-
small distribution system transformers (from vided as independent, fast-acting short-circuit
100 kVA) up to large transformers of several hun- protection in addition to the Buchholz protection.
dred MVA. Differential protection offers fast, sel-
ective short-circuit protection, alone or as a sup-
plement to Buchholz protection. It is part of the
standard equipment in larger units from about
5 MVA.
2.1 Differential protection

Transformer differential protection contains a
number of additional functions (matching to
transformation ratio and vector group, restraint
against inrush currents and overexcitation).
Therefore it requires some fundamental consider-
ation for configuration and selection of the setting
values.
The additional functions integrated per relay are
advantageous. However, backup protection func-
tions have to be arranged in separate hardware
(other relay) for redundancy reasons. Therefore
the overcurrent-time protection contained in the Fig. 2 Protection of a three-winding transformer

2.2 Earth-fault differential protection 2.4 Integration of Buchholz protection

Earth-fault differential protection detects earth- The Buchholz protection of the transformer
faults in transformers in which the star (neutral) evaluates the gas pressure of the transformer tank
point has low resistance or is solidly earthed. It en- and therefore detects internal transformer faults
ables fast, selective disconnection in the event of quickly and sensitively. The following consider-
an earth fault in the winding. The protection is ations are necessary for integration:
based on a comparison of the star-point current
n The trip command of the Buchholz protection
ISP with the phase currents of the main winding.
should act on the circuit-breaker directly and
independently of the differential protection
should be recorded in the fault log/fault record

of the differential protection
Coupling the trip command via a binary input of
the differential protection provides informative
data for evaluation in the case of a fault.
Fig. 3 Connection of an earth-fault differential

protection relay
The pickup sensitivity should be ≤ 10 % of the

current in the event of terminal earth fault (90 %
protected zone). The single-phase auxiliary mea-
suring input with connection IZ1 of 7UT613
should be used for this and assigned to the corre-
sponding main winding by setting. The earth cur-
rent of this input is then compared with the phase
currents of the main winding.
Fig. 4 Scheme for Buchholz protection
2.3 Backup protection functions
The integrated overcurrent-time protection
(ANSI 51) in 7UT613 serves as backup protection
for faults in the system to which power is sup-
plied. Separate overcurrent protection on the
low-voltage (LV) side is therefore unnecessary.
The 7SJ602 relay can be used as backup protection
against short-circuits in the transformer and as
additional backup protection against faults on the
LV side. The high-set, fast tripping stage I>>
(ANSI 50) must be set above the through-fault
current, so that it does not pick up in case of faults
on the low-voltage (LV) side.
The delayed trip (ANSI 51) must be of higher pri-
ority than the overcurrent protection in 7UT613.
Owing to the different ratings, windings S2 and S3
are assigned a separate overload protection (inte-
grated in 7UT613). The delta winding (often only
used for own internal supply) has its own overcur-
rent-time protection (ANSI 51, integrated in
7UT613) against phase faults.
At low ratings of the tertiary winding and accord-
ingly adapted transformer ratio, it should be
checked whether an external matching trans-
former may be required.

n 3. Settings The relay requires the following data for the pri-
3.1 Setting instructions for differential protection mary winding (side S1):
The differential protection as a main function of n The primary rated voltage UN in kV
the 7UT613 is parameterized and set in a few (line-to-line)
steps: n The rated apparent power
n Parameterize “three-phase transformer” pro- n The conditioning of the star point
tected object n The transformer vector group
n Assign the measuring locations on the main
Generally, the currents measured on the second-
protected object
ary side of the current transformers with a current
Example: flowing through them are not equal. They are de-
termined by the transformation ratio and the vec-
tor group of the transformer to be protected, and
by the rated currents of the current transformers.
The currents therefore have to be matched first to
make them comparable.
This matching takes place arithmetically in the
7UT613. External matching equipment is there-
fore normally not necessary. The digitized cur-
rents are converted to the respective transformer
rated currents. To do this, the transformer’s rating
data, i.e. rated apparent power, rated voltage and
the primary rated currents of the current trans-
formers, are entered in the protection relay.
Fig. 5 Assigning of measurement locations
Sides:
S1 HV side of the main protected object
(transformer)
73 MVA 72 MVA
S2 LV side of the main protected object INSide 1 = = 378 A I N Side2 = = 1663 A
(transformer) 3 ⋅ 110 kV 3 ⋅ 25 kV
S3 Side of the tertiary winding of the main protected I N Obj = I N Side I N Obj = I N Side
object (transformer) 400 A 2000 A
k1 = k2 =
Measuring locations assigned 3-phase: 378 A 1663 A
M1 Measuring location assigned to the main protected 16 MVA
object for side 1 I N Side3 = = 924 A
3 ⋅ 10 kV
M2 Measuring location assigned to the main protected
72 MVA
object for side 2 I N Obj = = 4157 A
M3 Measuring location assigned to the main protected 3 ⋅ 10 kV
object for side 3 1000 A
k3 =
When defining the sides, the assignments made 4157 A
regarding the measuring locations (Fig. 5) at the
Fig. 6 Magnitude matching
main protected object must be observed. Side 1 is
always the reference winding and therefore has
current phase position 0° and no vector group Fig. 6 shows an example of magnitude matching.
code. This is usually the HV winding of the trans- The primary rated currents of the two sides
former. The object data refer to specifications for (windings) S1 (378 A) and S2 (1663 A) are calcu-
every side of the protected object as fixed in the lated from the rated apparent power of the trans-
assignment definition. former (72 MVA) and the rated voltages of the
windings (110 kV and 25 kV). Since the current
transformer's rated currents deviate from rated
currents of these sides, the secondary currents are
multiplied by the factors k1 and k2.

The third winding (S3) on the other hand is only system (double earth fault in a non-earthed sys-
dimensioned for 16 MVA (e.g. as auxiliary supply tem). Thus, zero-sequence currents are of no con-
winding). The rated current of this winding (= cern for the stability of the differential protection
side of the protected object) is therefore 924 A. as they cannot occur in case of external faults. In
For the differential protection, however, compara- the case of internal faults, on the other hand, the
ble currents must be used for the calculation. zero-sequence currents (because they come from
Therefore, the protected object rated power of the outside) are absorbed almost totally by the
72 MVA must also be used as a basis for the third sensitivity. A very high sensitivity in the event of
winding. This results in a rated current (here: cur- earth faults in the protected zone can be achieved
rent under rated conditions of the protected ob- with overcurrent-time protection for zero-
ject, i.e. at 72 MVA) of 4157 A. This is the refer- sequence current and/or the single-phase overcur-
ence variable for the currents of the third winding. rent-time protection, which can also be used as
high impedance differential protection. The dif-
The currents are therefore multiplied by factor k3.
ferential protection function must be activated by
The relay does perform this matching on the basis
parameterization. The differential protection relay
of the set rated values. Together with the vector
7UT613 is delivered in inactive-circuit state. This
group which also has to be entered, it is able to
is because the protection may not be operated
perform the current comparison according to
without at least having set the vector groups and
fixed arithmetic rules. This is explained by the
matching values correctly first. The relay may re-
following example for the vector group Y(N)d5
act unpredictably without these settings.
(with earthed star-point):
Setting of the characteristic of the differential pro-
tection is based on the following considerations:
n The presetting of 0.2 x IN referred to the rated
current of the transformer can be taken as a
pickup value for the differential current as a rule.
n The slope 1 together with base point 1 take into
account current-proportional error currents
which may be caused by transformation errors of
the CTs. The slope (gradient) of this section of
the characteristic is set to 25 %.
n The add-on restraint increases the stability of
the differential protection in the very high
short-circuit current range in the event of exter-
nal faults; it is based on the setting value EXF-
restraint (address 1261) and has the slope 1 (ad-
dress 1241).
n The slope 2 together with base point 2 lead to
Fig. 7 Phasor diagram for vector group matching higher stabilization in the higher current range
at which current transformer saturation can oc-
Fig. 7 shows the windings and below them the cur. The slope of this section of the characteris-
phasor diagrams of symmetrical currents. The tic is set to 50 %.
matrix equation in a general form is:
( I m ) = k ⋅ ( k) ⋅ ( I n )
The phase currents on the left-hand (star-point)
side are equal to the winding currents (The mag-
nitude matching of the absolute value is not taken
into account in the figure).
Since there is no point earthed within the pro-
tected zone, no considerable zero-sequence cur-
rent (residual current) can be produced within the
protected zone in case of an earth fault outside the
protected zone. This is also valid if the system
star-point is earthed anywhere else in the system.
In case of an earth fault within the protected zone,
a zero-sequence current may occur at a measuring
location if the system star-point is earthed any-
Fig. 8 Tripping characteristics of the differential protection
where else or another earth fault is present in the

3.1.1 Notes on add-on restraint 3.1.3 Notes on setting the overexcitation blocking
In systems with very high through-flowing cur- Stationary overexcitation in transformers is char-
rents, a dynamic add-on restraint (stabilization) acterized by odd harmonics. The third or fifth
becomes effective for external faults. Note, that harmonic is suitable for stabilization. Since the
the restraint current is the arithmetic sum of the third harmonic is often eliminated in transform-
currents flowing into the protected object, i.e. is ers (e.g. in a delta winding), the 5th harmonic is
twice the through-flowing current. The add-on re- mostly used. The proportion of 5th harmonics
straint does not affect the I>> stage. which leads to blocking of the differential protec-
tion is set at 30 % (default). It is usually not neces-
The maximum duration of add-on restraint after
sary to set the crossblock function in this case.
detecting an external fault is set in multiples of a
period (AC cycle). The recommended setting
3.2 Earth-fault differential protection
value is 15 periods (preset). The add-on restraint
is disabled automatically – even before the set time The earth-fault differential protection detects –
period expires – as soon as the relay has detected selectively and with high sensitivity – earth faults
that the operation point IDiff/IRest is located steadily in transformers with earthed star-point. The pre-
(i.e. for at least one period) within the tripping requisite is that a current transformer is installed
zone. The add-on restraint operates separately for in the star-point connection, i.e. between the star
each phase. Hence, blocking can be extended to all point and the earthing electrode. This star-point
three phases thanks to the available vector-group transformer and the phase current transformer
(so called “crossblock” function). The recommen- define the limits of the protected zone.
ded setting value for the crossblock function is 15 No current ISt flows in the star-point connection
periods (preset). in normal operation. The sum of the phase cur-
rents
3.1.2 Notes on setting the inrush blocking
3I0 = IL1 + IL2 + IL3 is almost zero.
An inrush current with a high proportion of 2nd
harmonics is generated when switching on the In the event of an earth fault in the protected zone
transformer, which can lead to false tripping of (Fig. 9) a star-point current ISt will flow; depend-
the differential protection. The default for the ing on the earthing conditions of the system, an
harmonic restraint with 2nd harmonics of 15 % earth current can also feed the fault position via
can be retained without change. A lower value can the phase current transformer (dotted arrow),
be set for greater stabilization in exceptional cases which, however, is more or less in phase with the
under unfavorable energizing conditions resulting star-point current. The currents flowing into the
from the design of the transformer. protected object are defined positive.
The inrush restraint can be extended by the
“crossblock” function. This means, that all three
phases of the IDiff> stage are blocked when the
harmonic component is exceeded in only one
phase. A setting value of 3 periods, effective for the
time of mutual blocking after exceeding the differ-
ential current threshold, is recommended (de-
fault).
Fig. 9 Currents in case of an earth fault inside the

transformer

In the event of an external earth fault a zero- The earth-fault differential protection function
sequence current also flows through the phase cur- must be activated by parameterization. The
rent transformers. This current has the same mag- 7UT613 earth-fault differential protection relay is
nitude as the star-point current on the primary delivered in inactive-circuit state. This is because
side and is phase-opposed to it. Therefore, both the the protection relay may not be operated without
magnitude of the currents and their relative phase at least having set the allocation and polarity of
positions are evaluated for stabilization. This pro- the current transformers correctly first. The relay
duces the following tripping characteristic for the may react unpredictably without these settings.
earth differential protection:
The setting of the I-EDS> value is decisive for the
sensitivity of the protection. This value is the
earth-fault current which flows through the star-
point connection of the transformer. Another
earth current flowing in from the system is not
absorbed by the pickup sensitivity. The current
value refers to the rated operating current of the
side of the transformer to be protected. The pre-
set pickup value of 0.15 I/InS is normally appro-
priate.
3.3 Backup protection functions

3.3.1 Overcurrent-time protection
The definite-time overcurrent-time protection of
the 7UT613 serves as backup for the short-circuit
protection of the downstream system sections
when faults cannot be cleared in time there,
meaning that the protected object is in danger.
The overcurrent-time protection can be assigned
to one of the three voltage sides of the trans-
Fig. 10 Tripping charac-
teristic for earth In the above examples it was assumed that 3I0” and former. Correct allocation between the measur-
differential pro- 3I0’ are phase-opposed in the event of an external ing inputs of the relay and the measuring loca-
tection earth fault, which is correct as far as the primary tions (current transformer sets) of the power
variables are concerned. However, current trans- plant must also be observed. The stage I>> to-
former saturation can simulate a phase shift be- gether with stage I> or stage IP produces a two-
tween the star-point current and the sum of the stage characteristic. If the overcurrent-time pro-
phase currents, which weakens the restraint value. tection acts on the feed side of the transformer,
The restraint is zero at ϕ(3I0”; 3 I0’) = 90°. This cor - stage I>> is set so that it picks up for short-cir-
responds to the conventional method of direction cuits extending into the protected object, but not
detection by use of the vectorial sum and differ- for a short-circuit current flowing through it.
ence comparison.
The following phasor diagram shows the restraint
value in the event of an external fault:
Fig. 11 Phasor diagram of
restraint (stabiliza-
tion) value

Calculation example: 3.3.2 Load unbalance protection (negative-

Transformer Y(N)d5 sequence protection)
72 MVA The load unbalance protection (phase-balance
25 kV/10 kV current protection or negative-sequence protec-
uSC = 12 % tion) can be used in the transformer as sensitive
Current transformer 2000 A/1 A on the 25-kV-side protection function on the feed side for weak-
The overcurrent-time protection acts on the current single and two-pole faults. LV side, single-
25 kV side (= feed side). pole faults can also be detected which cause no
The maximum possible three-phase short-circuit zero-sequence current on the HV side (e.g. in vec-
current on the 10 kV side with rigid voltage on the tor group DYN).
25 kV side would be:
The load unbalance protection of the HV winding
1 1 SN Transf. (110 kV in the example) can detect the following
I 3polemax = ⋅ I N Transf. = ⋅ =
U SC Transf. U SC Transf. 3 ⋅U N fault currents on the LV side (25 kV in the
example):
1 72 MVA
⋅ = 13856.4 A
012
. 3 ⋅ 25 kV If I2 > = 0.1 A is set for the HV side, a fault
current of
110 kV 400 A
With a safety factor of 20 % this gives the primary I F1= 3 ⋅ ⋅ ⋅ 01
. A = 528 A
setting: 25 kV 1 A
I>> = 1.2 x 13856.4 A = 16628 A can be detected for a single-phase fault and of
With parameterization in secondary variables the 100 kV 400 A
I F2 = 3 ⋅ ⋅ ⋅ 01
. A = 305 A
currents in amperes are converted to the second- 25 kV 1 A
ary side of the current transformers.
for a two-phase fault on the LV side. This corre-
Secondary setting value: sponds to 26 % or respectively 15 % of the trans-
16628 A former’s rated current. Since this is a LV side
I >> = ⋅1 A = 8,314 A short-circuit, the delay time must be coordinated
2000 A
with the times of the subordinate protection relays.
i.e. at short-circuit currents above 16628 A (pri- The definite-time characteristic is two-stage.
mary) or 8,314 A (secondary), there is definitely a
short-circuit in the transformer area. This can be
cleared immediately by the overcurrent-time pro-
tection. Increased inrush currents are disarmed by
the delay times of the I>> stage if their fundamen-
tal exceeds the setting value. The inrush restraint
does not affect the stages I>>.
Stage I> represents the backup protection for the
subordinate busbar. It is set higher than the sum
of the rated outgoing currents. Pickup by overload
must be ruled out because the relay operates with
correspondingly short command times as
short-circuit protection in this mode and not as
overload protection. This value must be converted Fig. 12 Tripping characteristic of load unbalance protection
to the HV side of the transformer. The delay time
depends on the grading time in the outgoing lines. Stage I2> can be used for warning. A trip com-
It should be set 300 ms more than the highest mand can be given at the end of the delay time of
grading time on the LV side. Moreover, the inrush stage I2>>.
restraint for the I> stage must be parameterized
effectively in this case, so that false pickup of the
I> stage (resulting from the inrush of the trans-
former) is prevented.

3.3.3 Overload protection Time constant τ in thermal replica:

The thermal overload protection prevents over- The heating time constant th for the thermal rep-
load of the transformer to be protected. Two lica must be specified by the transformer manu-
methods of overload detection are available in the facturer. It must be ensured that the time constant
7UT6: is set in minutes. There are often other specifica-
tions from which the time constant can be deter-
n Overload protection with thermal replica mined:
according to IEC 60255-8,
n Hot-spot calculation with determining of the Example:
relative ageing rate according to IEC 60354. t6 time: This is the time in seconds for which 6
times the rated current of the transformer winding
One of these two methods can be selected. The may flow.
first is notable for easy handling and a low num-
ber of setting values; the second requires some τ th
= 0.6 ⋅ t 6
knowledge of the protected object, its ambient min
context and its cooling, and needs the input of the
If the transformer winding has a 6 time of 12 s
coolant temperature via a connected thermobox.
The second method is used when the transformer τ th
= 0.6 ⋅12 s = 7.2
is operated at the limit of its performance and the min
relative ageing rate is to be monitored by the hot-
spot calculation. the time constant must be set to 7.2 min.
Overload protection with thermal replica (to act
on the HV side) is chosen for this application 3.3.4 Overexcitation protection
example. Since the cause of the overload is nor- The overexcitation protection serves to detect in-
mally outside the protected object, the overload creased induction in generators and transformers,
current is a through-flowing current. The relay especially in power station unit transformers. An
calculates the temperature rise according to a increase in the induction above the rated value
thermal single-body model by means of the ther- quickly leads to saturation of the iron core and
mal differential equation high eddy current losses which in turn lead to im-
permissible heating up of the iron.
2
dΘ 1 1  I 
+ ⋅Θ = ⋅  Use of the overexcitation protection assumes that
dt τ th τ th  I N Obj  measuring voltages are connected to the relay. The
overexcitation protection measures the voltage/
The protection function therefore represents a frequency quotient U/f, which is proportional to
thermal replica of the object to be protected the induction B at given dimensions of the iron
(overload protection with memory function). core. If the quotient U/f is set in relation to voltage
Both the history of an overload and the heat trans- and frequency under rated conditions of the pro-
mitted to the ambient area are taken into account. tected object UNObj/fN, a direct measure is ob-
Pickup of the overload protection is output as a tained of the induction related to the induction
message. under rated conditions B/BNObj. All constant vari-
Notes on the setting ables cancel each other:
In transformers, the rated current of the winding U
to be protected, which the relay calculates from U N Obj
B U/f
the set rated apparent power and the rated voltage, = =
BN Obj f U N Obj / f N
is significant. The rated current of the side of the
main protected object assigned to the overload fN
protection is used as the basic current for detect-
The relative relation makes all conversions unnec-
ing the overload. The setting factor k is deter-
essary. All values can be specified directly related
mined by the ratio of the thermally permissible
to the permissible induction. The rated variables
continuous current to this rated current:
of the protected object have already been entered
I max in the 7UT613 relay with the object and trans-
k=
I N Obj former data when setting the differential protec-
tion.
The permissible continuous current is at the same
time the current at which the e-function of the
overtemperature has its asymptote. The pre-
setting of 1.15 can be accepted for the HV wind-
ing.

Setting instructions
The limit value of permanently permissible induc-
tion in relation to the rated induction (B/BN)
specified by the protected object manufacturer
forms the basis for setting the limit value. This
value is at the same time a warning stage and the
minimum value for the thermal characteristic
(see Fig. 13).
Fig. 14 Integration into substation control system
Via the interface,

n Messages
n Alarms
n Measured values
are transmitted from the transformer differential
protection to the substation control system. Mes-
sages are available for every one of the activated
Fig. 13 Tripping characteristic of the overexcitation protection protection functions, which can be either trans-
mitted to the substation control system in the
An alarm is output after the relevant set delay time course of plant equipment parameterization, or
(about 10 s) of the overexcitation stage U/f has ex- allocated to the LEDs or alarm contacts in the
pired. Major overexcitation endangers the pro- protection relay. This configuration is made clear
tected object already after a short time. The high- and easy by the DIGSI matrix.
set tripping stage U/f>> is therefore set to a maxi-
mum of 1 s. Service interface
The thermal characteristic should simulate the The 7UT613 has a separate service interface which
heating, i.e. temperature rise, of the iron core re- can be read out by telecommunication via a mo-
sulting from overexcitation. The heating charac- dem. The user is informed in the office quickly
teristic is approximated by entering 8 delay times and in detail about the transformer fault. The data
for 8 given induction values B/BNObj (referred to in are then analyzed in the office by DIGSI. If this
simplified form as U/f). Intermediate values are remote fault clearing is insufficient, the fault data
obtained by linear interpolation. If no data are provide hints for an efficient service mission.
available from the protected object manufacturer,
the preset standard characteristic is used.
n 4. Further functions
4.1 Integration into substation control system
The protection can be connected to a substation
control system via the system interface and oper-
ated in parallel by PC via the service interface to a
star coupler for separate remote communication.

n 5. Connection diagram n 6. Summary

Optimum protection of the transformer with
SIPROTEC relays means security of investment
for valuable operating equipment and therefore
makes a contribution to maximum supply secu-
rity.
From a technical point of view, the SIPROTEC
7UT613 relay offers extensive short-circuit pro-
tection for both the main and the backup protec-
tion of transformers in one single relay. Extensive
measuring functions allow trouble-free connec-
tion of the relay without the need for additional
equipment, and enable monitoring of the trans-
former in operation in terms of its electrical and
thermal parameters. The relay’s presettings are
chosen so that the user need only parameterize
the known data of the main and primary trans-
formers. Many of the default settings can simply
be accepted as they are, and therefore make
parameterization and setting easier.
Fig. 15 Connection diagram of 7UT613

Protection of a Trans-
former with Tap Changer
n 1. Introduction
Transformers are among the most important and
cost-intensive equipment in electrical power sys-
tems, meaning that faults which occur in these
components not only entail an interruption in the
electrical power supply over wide areas but also
cause considerable losses in financial terms. A con-
tinuous fault-free power supply must therefore be
ensured, over the course of years if possible. Faults
LSP2171.eps
and signs of potential failures of the transformers
must therefore be detected in time in order to take
suitable measures for troubleshooting.
For this reason transformers are equipped with
various monitoring and protection relays depend-
ing on their type and size. The electrical protec- Fig. 1 SIPROTEC 7UT6 transformer protection
tion should be highlighted particularly in addition
to the mechanical protection. n 2. Protection concept
Fuses and definite-time overcurrent-time relays Depending on the type and size of the transform-
are sufficient in smaller distribution transformers ers, Buchholz protection, overload protection and
for both technical and economic reasons. Fuses overcurrent time protection are used as fast, selec-
and definite-time overcurrent-time relays repre- tive short-circuit protection in addition to the
sent time-delayed protection measures. Time-de- classic differential protection (as from approxi-
layed protection tripping relays are unacceptable mately 1 MVA and higher). These are only men-
for larger transformers in distribution, transmis- tioned briefly here because they are described in
sion and power generation applications and must detail in other application examples.
be disconnected immediately to avoid system in-
stability and cost-intensive shutdowns. 2.1 Differential protection as main protection
Differential protection represents the main pro-
Transformer faults can generally be divided into
tection function for the transformer and is fea-
five categories:
tured in the SIPROTEC relays 7UT6* (addr. 1201)
n Interturn and terminal fault and 7UM62* (addr. 2001). It also comprises a
n Winding fault number of additional functions (matching to
n Fault on the transformer tank and auxiliary transformation ratio and vector group, restraint
devices against inrush currents). Therefore false differen-
n Fault on the transformer tap changer tial currents caused by transformation errors of
the current transformers are to be expected in
n Abnormal operating conditions (temperature,
practice. In regulated transformers an additional
humidity, dirt)
error current is to be expected caused by adjust-
n External fault ment of the tap changer.
This application example gives an insight into the The additional functions integrated in the relays
protection of regulated power transformers with are influenced by the use of a transformer with tap
tap changer function. changer and the resulting correction values. This
is explained in chapter 4 by a calculation example.

Relay characteristic
∆IGF = Total error current
∆IWF = CT error current

∆IAF = Matching error current
∆IWF = CT magnetization
current
Fig. 2 False differential current in the event of load and Fig. 4 Earth-fault differential protection
continuity faults and matched relay characteristic
Backup protection such as overcurrent time 2.3 Buchholz protection

protection are provided in separate relays (e.g. The Buchholz protection is coupled into the relay
7SJ602, 7SJ45/46). The overcurrent-time protec- (alarm, tank and trip message) as an external pro-
tion and/or overload protection contained in the tection (7UT6*, 7UM62* addr. 8601, 8701) and is
differential protection relays serve merely as used for liquid-cooled transformers and reactors
backup protection against external faults in the with expansion tank. The Buchholz relay responds
connected power system. to faults which cause the forming of gas in the
tank (winding fault, interturn fault, loss of insulat-
ing fluid, accumulation of air).
87T Transformer differential
protection
n 3. Integration of the transformer tap changer
51 Assigned to W2 as
protection for busbar in differential protection
faults and backup pro- 3.1 Purpose of a transformer tap changer
tection for feeders
Voltage regulation on transformers with load tap
50/51 Backup protection for
the transformer changers is an important topic for power supply
49 Overload protection companies. In accordance with DIN/IEC standards
Bu Buchholz protection it is necessary to keep the 230 V/400 V voltage in
the public low-voltage system constant, at least in
the range ± 10 %. To keep the voltage constant in
this bandwidth, a transformer tap changer is con-
trolled by a transformer voltage regulator (e.g.
Maschinenfabrik Rheinhausen TAPCON®
230/240). The voltage regulator constantly com-
pares the actual value Uact (output voltage at the
transformer) and a fixed or load-dependent
setpoint Usetp.
Fig. 3 Protection of a two-winding transformer
The voltage regulator supplies the setting variable
2.2 Earth-fault differential protection for the transformer’s load tap changer dependent
In transformer windings with star-point earthing on the deviation of the actual value from the
via an impedance (earth-current limiting), the setpoint. The load tap changer switches when the
earth-current differential protection (7UT6* addr. given bandwidth (Usetp +/- B%) is dropped below
1301) is an ideal supplement to the phase protec- or exceeded. The voltage at the transformer is thus
tion to enhance the response sensitivity in the kept constant. Fluctuations within the permissible
earth fault. bandwidth have no influence on the control be-
havior or the switching process.
In this method the measured star-point current
I0* in the transformer star-point is compared with
the calculated summation current I0** of the
phase currents.

The parameters of the voltage regulator can be system data. Former relay generations required
adapted optimally to the behavior of the system separate matching transformers for (e.g.) vector
voltage so that a balanced control behavior is group adaptation.
achieved at a low number of cycles of the load tap
changer.
Fig. 5 Voltage regulation of a regulated transformer using a TAPCON® system
3.2 Correcting “false” differential currents If the winding is regulated, not the actual rated
Most calculations of differential and restraint cur- voltage is used as UN for the stabilized side, but the
rents are made without taking the tap changer po- voltage corresponding to the mean current of the
sition into account. In practice, however, most regulated range.
power transformers are equipped with a tap U max ⋅U min 2
changer. Two types are distinguished: UN = 2⋅ =
U max + U min 1
+
1
n Off-load tap changer
U max U min
n On-load tap changer
Whilst most transformers are equipped for with Umax, Umin as limits for the regulated range.
off-load tap changing, on-load tap changing is Example:
used for voltage regulation in power systems. The
protection parameterization must take the differ- Transformer Ynd5
ent tap changer positions into consideration to 35 MVA
avoid the possibility of false tripping (especially 110 kV/20 kV
with extreme positions). Y side regulated ± 20%
Correct operation of the differential protection re- For the regulated winding (110 kV) this results in
quires that the differential currents on the primary Maximum voltage Umax = 132 kV
and secondary side correspond to real conditions Minimum voltage Umin = 88 kV
under normal load and fault conditions. The pri-
mary and secondary side current transformers do Voltage to be set
not pick up the real transformer ratio. Today’s 2
protection relays such as those in the SIPROTEC U N -PRI SIDE 1 = = 105.6 kV
1 1
series compensate these faults with calculated cor- +
132 kV 88 kV
rection factors based on the parameterized power

Parameters in relevant SIPROTEC relays corresponds on the CT1 secondary side to

7UT612 addr. 240 I N1 402.3 A
IN 1 = = = 100575
. A ≅ I NObj
7UT613/63* addr. 311 CT1 400
7UM62 addr. 240
(referred to S1)
n 4. Calculation example Object rated current of the unregulated side
Influence of tap positions on differential (remains constant)
and restraint currents
A two-winding transformer with a tap change SN 72 MVA
I N2 = = = 1663 A
range of –15 % to + 5 % is used for the following 3 ⋅U N2 3 ⋅ 25 kV
example. The tap changer is integrated in the pri-
mary winding for voltage regulation. corresponds on the CT2 secondary side to
I N2 1663 A
I N2 = = = 0.8315 A ≅ I NObj
CT2 2000
(referred to S2)
4.2 Calculations of the differential/restraint

72 MVA 72 MVA
I N SIDE 1 = = 378 A I N SIDE 2 = = 1663 A currents in the tap changer extreme positions
3 ⋅ 110 kV 3 ⋅ 25 kV
4.2.1 Tap position +5 %
Object current in maximum tap position
Fig. 6
SN 72 MVA
I N 1 ( + 5 %) = = = 359.9 A
Transformer YNd5 3 ⋅U max 3 ⋅115.5 kV
(irrelevant for the calculation) corresponds on the CT1 secondary side to
72 MVA
110 kV/25 kV I N1 (+5 %)
359.9A
I N 1 ( + 5 %) = =
= 0.8997 A ≅ 0.8946 ⋅ I NObj
Y side regulated -15 %/+5 % CT1 400
CT1 = 400 (1 A) Differential current in maximum tap position
CT2 = 2000 (5 A)
I Diff = I N 1 ( +5%) − I NObj = 0.8946 ⋅ I NObj − I NObj = 01054
. ⋅ I NObj
4.1 Calculations of the voltage, object rated
currents and correction factors to be set Restraint current in maximum tap position
The voltage to be set is calculated using the for-
mula in chapter 3.2 and parameterized as I Restraint = I N1 ( +5%) + I NObj = 0.8946 ⋅ I NObj + I NObj = 18946
. ⋅ I NObj
UN WIND.S1 in the SIPROTEC relays 7UT6*,
7UM62*.
4.2.2 Tap position -15 %
For the regulated winding (110 kV) this gives a
Object current in minimum tap position
calculated
maximum voltage Umax = 115.5 kV SN 72 MVA
I N 1 (− 15 %) = = = 444.6 A
minimum voltage Umin = 93.5 kV 3 ⋅U min 3 ⋅ 935
. kV
Voltage to be set corresponds on the CT1 secondary side to
U ⋅U 2
U N 1 = 2 ⋅ max min = I N1 (-15 %)
444.6 A
U max + U min 1 1 I N 1 (− 15 %) = = = 11115
. A ≅ 11051
. ⋅ I NObj
+ CT1 400
U max U min
Differential current in maximum tap position
115.5 kV ⋅ 93.5 kV
= 2⋅ = 103.3 kV
115.5 kV + 935 . kV I Diff = I N 1 (−15%) − I N2 = 11051
. ⋅ I NObj − I NObj = 01051
. ⋅ I NObj
Restraint current in maximum tap position
Object rated current of the regulated side
SN 72 MVA I Restraint = I N 1 ( +5%) + I N2 = 1051
. ⋅ I NObj + I NObj = 21051
. ⋅ I NObj
IN 1 = = = 402.3 A
3 ⋅U N 1 3 ⋅103.3 kV

At the voltage to be set according to chapter 3.2

the same differential current portion of the object
rated current is measured respectively in the ex-
treme positions ï The calculated voltage UN1 to
be set corresponds to the middle position of the
transformer tap changer.
4.3 Difference between operating current and

restraint current
I op = m ⋅ I Restraint
Presetting m = 0.25
Iop = 0.25 ⋅ IRestraint
At maximum tap position + 5 % it follows that
I op = 0.25 ⋅18496
. ⋅ I NObj = 0.4624 ⋅ I NObj
At minimum tap position - 15 % it follows that
I op = 0.25 ⋅ 21051
. ⋅ I NObj = 0.5263 ⋅ I NObj
From the calculations it can also be derived that, Fig. 8 Tripping characteristic of the differential protection in 7UT6* and 7UM62*
under rated conditions and at the tap changer ex-
treme positions, the operating currents are not in The matching is currently performed by correc-
the tripping area (due to the characteristic). tion of the primary voltage according to the
Therefore the slope (gradient) of the trip charac- formula in chapter 3.2 and parameterization by
teristic (7UT6* addr. 1241, 7UM62* addr. 2041) means of the appropriate addresses or DIGSI.
need not be adapted to conditions (presetting
m = 0.25). n 6. Integration of tap positions in DIGSI
Transformer taps can be indicated either by the
n 5. Parameterization notes DIGSI PC or the graphic display of the SIPROTEC
Direct coupling of the transformer tap changer relay. The transformer taps are signaled via binary
into the protection algorithm is available as from inputs on the relay. The binary inputs are assigned
V4.6 for 7UT63* relays (approx. available as of according to the coding type and number of trans-
mid-2005). By reading in the tap positions (with former taps (see Fig. 9).
the codings BCD, binary, 1 from n table), the
transformation ratio can be adapted depending on
the position, and the faulty differential currents
compensated as a result. This improves both the
sensitivity and the stability of the differential pro-
tection.
LSP2669.tif
Fig. 9 Schematic diagram – Reading in of tap position by DIGSI or

graphic display
Fig. 7 7UT613/63* parameterization in transformer
system data
In order to display the transformer taps, the trans-
former tap message type must first be entered in
the configuration matrix.

LSP2671.tif
Fig. 12 Object properties, transformer tap message
n 7. Summary
In most differential protection relays the influence
of the transformer tap changer is mainly taken
into account with the corrected input of the pri-
LSP2670.tif
mary voltage (determined in the middle position
of the tap changer).
Fig. 10 Activating the transformer tap message in the
information catalog Owing to the demand for stabilized voltages and
regulation by tap changers, the probabilities of
The transformer tap message is entered in the tripping faults in the extreme positions of the tap
configuration matrix and activated by configura- changer can be limited in future by matching of
tion of the binary inputs. the transformation ratio. This requirement is met
by inclusion of the tap position in the protection
LSP2575.tif
relay and consideration of the protection algo-

rithm in the SIPROTEC 7UT63* relays.
Fig. 11 Entry of the transformer tap message in the
DIGSI configuration matrix
Detailed settings must be made under the object

properties of the transformer tap message.
n Number of bits: necessary for coding; number
depends on the selected coding
n Display offset: the value by which the size of the
displayed value is to be shifted in positive or
negative direction in relation to the magnitude
of the actual value
n Moving contact: if the moving contact option is
activated, the tap position is not recognized as
valid and accepted until the moving contact sig-
nals that it has reached the taps (always the
highest ranking contact)

Protection of an
Autotransformer
n 1. Introduction
Transformers are valuable equipment which make
a major contribution to the supply security of the
power system. Optimum design of the transform-
er protection ensures that any faults that may oc-
cur are found quickly, so that consequential
damage is minimized. A special variant is the
so-called autotransformer in which, unlike in the
full transformer, the voltage and current transfor-
mation is not performed by two independent
windings but uses part of the winding from both
sides, allowing a much more compact design.
The spectrum of autotransformers ranges from 7UT613
small distribution system transformers (from
1000 kVA) to large transformers of several hun- 7UT612
dred MVA. Their use becomes more interesting,
the less the ratio between the high-voltage (HV) Fig. 1 SIPROTEC 7UT6 transformer protection
side and low-voltage (LV) side deviates from 1,
i.e. the less energy is transmitted via the magnetic
coupling which leads to a saving in iron material.
In addition to the design notes, a complete setting
example with SIPROTEC protection relays for a
triple-wound autotransformer in the transmission
system is described.
Differential protection offers fast, selective short-
circuit protection, alone or as a supplement to the
Buchholz protection. In larger units from about
5 MVA it is part of the standard equipment. In ad-
dition to the main protection function which reli-
ably clears a short-circuit within the protected ob-
ject, a fully-fledged protection concept contains a
number of additional functions which take care of
other problems such as overload, overexcitation,
etc. All the necessary functions are already con-
tained in the SIPROTEC 4 relays. Backup protec-
tion functions are a useful addition.
Fig. 2 shows an example of a full protection
concept for an autotransformer.
Fig. 2 Protection of an autotransformer

For this example a differential protection n 3. Structure of an autotransformer

ANSI 87T (3-windings version) must be selected
(7UT613), so that the delta stabilizing winding
can be included in the protected zone (different WI WII
I1L1
alternatives for connection are explained in detail U1L1 I2L1
in chapter 3). If this is missing, an autotransfor- I1L2 U2L1
mer can also be fully protected with the 7UT612. U1L2 I2L2
The integrated overload protection ANSI 49 is I1L3 U2L2
adapted to the load and protects the transformer U1L3 I2L3
from overheating and premature aging. The over- U2L3
excitation protection ANSI 24 prevents impermis-
sible heating of the iron.
To increase the earth-fault sensitivity, an addition-
al high-impedance earth-fault differential W I + W II U1 I 1
N= = N; 1 =
protection ANSI 87TE is often used in English- W II U2 I2 N
influenced regions. The single-phase relay
Fig. 3 Structure and transmission ratios of a
7VH600 is on the parallel circuit of the outgoing two-winding autotransformer
side transformers and of the star-point transform-
er. The seven transformers (3 side 1, 3 side 2,
Autotransformers only have the vector group
1 star point) must be provided additionally, how-
YNyn0; i.e. there is no phase shift of the currents
ever, and be designed according to class TPS
and voltages between the primary and secondary
(IEC 60 044-6). A pickup value of 10 % In is usual-
sides. The common star point is always earthed.
ly achieved. Additionally or alternatively, an over-
Therefore both sides are always electrically con-
current relay Ie>t ANSI 50N can be provided in
nected. The distribution of the star-point current
the star-point link. However, this must be in time
on both sides depends on several factors, e.g. win-
coordination with the subordinate overcurrent re-
ding distribution and presence of a tertiary win-
lay.
ding.
The delta winding which, in addition to its stabi-
lizing (restraint) function, is also often used for
n 4. Implementation with SIPROTEC
autonomous supply, gets its own overcurrent-
time protection ANSI 50, 50N for external phase 4.1 Differential protection
faults. The voltage relay 7RW600 (ANSI 59N) on Transformer differential protection contains a
the open delta winding of the voltage transformer number of additional functions (matching to
measures the displacement voltage 3U0 with transformation ratio and vector group, stabilizati-
which an earth fault in the tertiary winding or in on (restraint) against inrush currents and over-
the connected distribution system is indicated. excitation) and therefore requires some funda-
An overcurrent-time protection ANSI 50, 50N is mental consideration for configuration and
arranged on the HV and LV sides, each with a selection of the setting values. The additional
instantaneous tripping stage I>> and delayed functions integrated per relay can be used to ad-
stage I> (against phase and earth-faults). Integrat- vantage. However, it must be considered that ba-
ed overcurrent-time protection can also configur- ckup protection functions must be arranged in
ed optionally in 7UT613 on one of the two sides. separate hardware (further relay) for reasons of
For every outgoing circuit the breaker failure pro- hardware redundancy. This means that the over-
tection ANSI 50 BF must be activated in the rele- current-time protection in the differential protec-
vant protection relay. tion 7UT612/ 613 can only be used as backup
protection against external faults in the connected
The individual elements shown in an overview are power system. The backup protection for the
described step by step as follows. transformer itself must be provided as a separate
overcurrent relay (e.g. 7SJ602). The Buchholz
protection as fast short-circuit protection is supp-
lied with the transformer.

Designations according to ANSI (American Na-

tional Standard) are used for the individual Phase L1 Phase L2 Phase L2 Sides:
functions. The differential protection therefore S1 HV side of the autotransformer
winding on the main protected
has ANSI No. 87, for example. The differential object
protection is provided as a definite-time fast S2 LV side of the autotransformer
winding (tapping) on the main
short-circuit protection in addition to the Buch- protected object
holz protection. S3 Side of the tertiary winding
(stabilizing winding) on the main
The differential protection for an autotransformer protected object
can be implemented in 2 different ways depending 3-phase measuring locations, assigned:
on the available current transformers. M1 Measuring location for side 1
assigned to the main protected
object
1. Differential protection over the whole trans- M2 Measuring location for side 2
former bank (protection to be used 7UT612 assigned to the main protected
object
(2 windings) / 7UT613 (3 windings): M3 Measuring location for side 3
In this case, as shown in Fig. 4, three phase current assigned to the main protected
object
transformers are used for each side. The star-point Additional measuring locations 1-phase,
transformer is insignificant for the differential assigned (summation current of the
transformer set):
protection but can be used for backup overcur- Z3 Measuring location for side 1 and
rent-time protection. It can be set on both sides side 2 assigned to the main
protected object
depending on further use by other protection
functions.
2. Current comparison per autotransformer
winding
Use of phase current transformers upstream of the
star-point junction (earth-current winding). Fig. 5 Fig. 4 Topology of a transformer bank consisting of 3 single-phase autotransformers
shows, that the supply lines to the star (neutral) with a stabilizing winding designed as a tertiary winding
point all have a phase current transformer. In this
case, the autotransformer can be treated as a node
object with three terminals.
Phase L1 Phase L2 Phase L2 Sides:
Both connection types are basically different and S1 HV side of the autotransformer
are treated separately in chapter 5. winding on the main protected
object
S2 LV side of the autotransformer
4.2 Earth-fault differential protection winding (tapping) on the main
protected object
The earth-fault differential protection cannot be S3 Star-point side of the autotrans-
used in the autotransformer. former on the main protected
object
3-phase measuring locations, assigned:
4.3 Backup protection functions M1 Measuring location for side 1
The integrated overcurrent-time protection object
(ANSI 51) in the 7UT613 serves as backup protec- M2 Measuring location for side 2
tion for faults in the supplied system. Separate object
overcurrent protection on the LV side is therefore M3 Measuring location for side 3
unnecessary. The 7JS600 relay on the HV side can object
be used as backup protection against short-circuits Additional measuring locations 1-phase,
assigned (summation current of the
in the transformer and as additional backup protec- transformer set)
tion against LV side faults. The high-set, fast trip- Z3 Measuring location for side 1 and
side 2 assigned to the main
ping stage I>> (ANSI 50) must be set above the protected object
short-circuit current flowing through it, so that it
does not pick up in the event of faults on the LV
side. The delayed tripping (ANSI 51) must be gra-
ded with the overcurrent protection in the 7UT613.
Fig. 5 Topology of a transformer bank consisting of 3 single-phase autotransformers;
The windings S1 and S2 can be protected with the topology definitions for one current comparison protection per phase, i.e. there
is phase-selective current measurement at M3 before the star-point junction
integrated overload protection.

The delta winding – which is often only used for n 5. Settings

own, internal supply – has its own overcurrent- 5.1 Setting instructions for differential protection
time protection (ANSI 51) against phase faults over the whole transformer bank
and because of its reduced rated power it requires The differential protection as a main protection
a separate overload protection. Both can be reali- function of the 7UT612/613 is parameterized and
zed with a 7SJ600 as shown in Fig. 2, for example. set in a few steps:
4.4 Integration of Buchholz protection n Parameterize protected object “autotransformer”

The Buchholz protection of the transformer eval- n Assign the measuring locations on the main
uates the gas pressure of the transformer tank and protected object
therefore detects internal faults in the transformer
quickly and sensitively. The following should be
considered for the integration:
should act on the circuit-breaker directly and
independently of the differential protection
should be recorded in the fault event log / fault Main

protection
record of the differential protection object
Coupling the trip command via the binary input
of the differential protection provides informative
data for evaluation in the event of a fault.
Sides:
S1 HV side of the main protected object (autotransformer)
S2 LV side of the main protected object (autotransformer)
S3 Side of the tertiary winding (stabilizing winding) of the
main protected object (autotransformer)
3-phase measuring locations, assigned:
M1 Measuring location for side 1 assigned to the main
protected object
protected object
protected object
protected object
Fig. 6 Integration in the Buchholz protection Fig. 7 Topology of an autotransformer with a stabilizing
winding designed as a tertiary winding
The topology in the autotransformer is defined as

follows:
Side 1 is the first of the autotransformer windings;
the HV side is usually chosen here.
Side 2 is the second of the autotransformer
windings; the LV side is usually chosen here.
This may be followed by further tappings. If there
is a delta stabilizing winding, this should be as-
signed last.

The relay requires the following data of the pro- The technical data of the 7UT612/613 show a per-
tected object: missible ratio of 0.25 < k < 4 specified for phase
currents. This means that in the case of the
n The primary rated voltage UN in kV
winding 3, k3 has an impermissibly small ratio.
(line-to-line)
A matching transformer must be provided here to
n The secondary rated voltage UN in kV
reach the permissible range. It should be dimen-
(line-to-line) sioned so that the matching factor reaches just
n The rated apparent power which is the same for
above the minimum figure of 0.25. In this case it
both sides in the autotransformer. could therefore be assumed:
The “autotransformer” setting in the configura-
tion automatically defines that no vector group n > 0.25 / (400A / 5773.5A) = 3.6, e.g. n=4.
shift is performed (phase angle 0° between HV
and LV side) and the zero current elimination is
performed on both sides. UN= 400 kV 300 A
k1 = = 104
,
288.7 A
In transformers, the currents measured on the
secondary side of the current transformer when 400 A
k3 = = 0,07
current flows through are not generally equal, but 5773.5 A
are determined by the ratio of the transformer to
SN = 200 MVA
be protected and the rated currents of the current
UN = 20 kV
transformers. To make the currents comparable SN = 12 MVA
they therefore have to be matched first. This takes
place arithmetically in the 7UT613. External 600 A
UN = 220 kV k2 = = 114
,
matching equipment is therefore normally super- 524.9 A
fluous (frequent exception: tertiary winding with
low rated apparent power). The digitized currents
are converted to the respective transformer rated
SN 200 MVA
currents. To do this, the transformer rating data, I NT 1 = = = 288.7 A → I NCT1 = 300 A
3 ⋅UN 3 ⋅ 400 kV
i.e. rated apparent power, rated voltages and the
200 MVA
primary rated currents of the current transform- I NT2 = = 524.9 A → I NCT2 = 600 A
3 ⋅ 220 kV
ers, are entered in the protection relay.
12 MVA
I NT3 = = 346 A → I NCT3 = 400 A
Fig. 8 shows an example for magnitude matching. 3 ⋅ 20 kV
The primary rated currents of the two sides S1 IN obj1 = INT1 = 288.7 A
(288.7 A) and S2 (525 A) are calculated from IN obj2 = INT2 = 524.9 A
the rated apparent power of the transformer IN obj3 =
200 MVA
= 5773.5 A
(200 MVA) and the rated voltages of the windings 3 ⋅ 20 kV
(400 kV and 220 kV). Since the current transfor-
mer rated currents deviate from these transformer Fig. 8 Example of an autotransformer for magnitude
rated currents, the secondary currents are multi- matching
plied by the factors k1 and k2. The third winding
(S3) on the other hand is only dimensioned for Together with the autotransformer information
12 MVA (e.g. as an auxiliary supply winding). The the protection relay is now able to perform a cur-
rated current of this winding (= side of the pro- rent comparison. The principle is explained in the
tected object) is therefore 346 A. For the differen- following example (see Figs. 9 and 10):
tial protection, however, comparable currents
according to the ratio of the individual sides of the
transformer must be used for the calculation.
Therefore the rated power of the protected object
of 200 MVA must likewise be taken as basis for the
third winding. This gives a theoretical rated cur-
rent (here = current under rated conditions of the
protected object, i.e. at 200 MVA) of 5773.5 A.
This is the reference variable for the currents of
the third winding. The currents are therefore
multiplied by factor k3. The relay performs this
magnitude matching based on the set rated values.

Vector group: Autotransformer (YNyn0) ü = 3/2 Both sides are converted to a “virtual” side, based
(Transformation ratio) on which the current comparison is made. In case
Side 2: w2 = 2N Side 1: w1 = N of vector group 0 (as is always the case with the
autotransformer) this is done by a standard ma-
trix. The matrices used are given by the standard
matrix by subtracting the zero current from the
measurement (corresponds to 1/3 of the sum of
all three phase currents). This is necessary, becau-
se it is not possible to divide the star-point
current on both sides of the protected object.
Side 2: Side 1: 5.2 Setting instructions for differential protection

 I 2A   2 −1 −1  I 2 L 1   I1A   2 −1 −1  I1L 1  in current comparison per autotransformer
  1       1     winding
I 2 B = ⋅ −1 2 −1 ⋅ I 2 L 2 I1B = ⋅ −1 2 −1 ⋅  I1L 2 
  3       3    
 I 2C   −1 −1 2   I 2 L 3   I1C   −1 −1 2   I1L 3  If current transformers are available for each
I 2L1 = −IF I1L 1 = ü ⋅ 2 / 3 ⋅ I F = I F
phase on the supply lines to the star point (earth
winding), one node protection per phase can be
I 2 L2 = 0 I1L2 = 0
implemented. Current transformation ratios and
I 2 L3 = 0 I1L3 = 0
their tap changes have no effect, because the three
 I 2A   −2 ⋅ I F +0 +0   I1A   2 ⋅ I F +0 +0   2I F  entry points of the current are measured here and
  1     1   1
I 2B = ⋅ I1B = ⋅ −1I F −0 +0 =  − I F 
  3  I F +0 +0    3   3   form the end points of the Kirchoff node.
 I 2C   I F +0 +0   I1C   −1I F +0 −0   − I F  Such a current comparison is more sensitive for
earth-faults than normal differential protection
 2I F   − 2I F  (see Fig. 11). This is of interest, because these
 IDiff A   3   3   0  faults have the highest probability in transformer
   IF   IF   
 IDiff B  =  −  +   =  0  à No trip banks. Any stabilizing winding or tertiary winding
 IDiff   3   3   0 
 C I I must not be integrated in the protection in this
− F   F 
 3   3  application, even if it is connected externally and
Fig. 9 Calculation of the differential current in the autotransformer for an external fault
equipped with current transformers, because this
does not belong to the protected (phase-selective)
Vector group: Autotransformer (YNyn0) ü = 3/2 node.
(Transformation ratio)
Side 2: w2 = 2N Side 1: w1 = N
W = 1 = W1 + W2
275 kV
W2 = = 0.6875
400 kV
W1 = W - W2 = 1 - 0.6875 = 0.3125
Side 2: Side 1: Example:

 I 2A   2 −1 −1  I 2 L 1   I1A   2 −1 −1  I1L 1  Fault in winding W2 at 50 %
  1       1    
I 2 B = ⋅ −1 2 −1 ⋅  I 2 L 2  I1B = ⋅ −1 2 −1 ⋅  I1L 2  WF = 0.5 W2 = 0.34375
  3       3    
 I 2C   −1 −1 2   I 2 L 3   I1C   −1 −1 2   I1L 3  I1 (W1 + WF) = I3 (W2- WF)
I 2L1 = 0 I1L 1 = ü ⋅ 2 / 3 ⋅ I F = I F W1 + WF
I3 = I1 = 1.909 ⋅ I 1
I 2 L2 = 0 I1L2 = 0 W2 − WF
I 2 L3 = 0 I1L2 = 0 Differential current in the
 I 2A   0  I1A   2 ⋅ I F +0 +0   2I F  2-winding differential protection
  1     1   1 acc. to Chap. 5.1
I 2B = ⋅ 0 = ⋅ − − + I 
 3  − F 

I 1 1 I 0 0 =
 3    B  3  F
IDiff = I1 (I2 = 0)
 I 2C   0  C
I 1  − 1 I F + 0 − 0  −
 I F 
Differential current in the
3-winding differential protection
 2I F   2I F  (incl. earth winding) acc. to
 IDiff   3   0   3  Chap. 5.2
 A  I     I 
 IDiff B  =  −  +  0  =  −  à Trip
F F
IDiff = I1 + I3 = I1 + 1.909 ⋅ I1
 IDiff   3   0   3 
 C I I = 2.909 ⋅ I1
− F  − F 
 3   3 
Fig. 11 Increasing the sensitivity by using
Fig. 10 Calculation of the differential current in the event of an internal fault phase-selective earth windings

The topology is determined as follows:

The two autotransformer windings become
S1 and S2 (tapping).
The earth winding must in this case be set as
side 3. If there is a further tapping, the earth wind-
ing is specified as side 4. As soon as one of the
sides is defined as earth winding, the protection
automatically places node differential protection
across all autotransformer windings involved.
The differential protection function must be
activated by parameterization. The differential
protection relay 7UT613 is delivered in inactive-
circuit state. This is because the protection may
not be operated without at least having set the
vector groups and matching values correctly first.
The relay may react unpredictably without these
settings.
Setting of the characteristic of the differential pro-
tection is based on the following considerations: Fig. 12 Differential protection tripping characteristics
n The presetting for IDiff > of 0.2 x IN referred to

the rated current of the transformer can be ta- add-on restraint does not affect the I>> stage.
ken as a pickup value for the differential current The maximum duration of add-on restraint after
as a rule. detecting an external fault is set in multiples of a
n The slope 1 together with base point 1 take into period (AC cycle). The recommended setting
account current-proportional false currents value is 15 periods (preset). The add-on restraint
which may be caused by transformation errors is automatically disabled before the set time peri-
of the CTs. The slope of this section of the cha- od expires as soon as the relay has detected that
racteristic is set to 25 %. the operation point IDiff/IRestraint is located steadily
n The add-on restraint increases the stability of (i.e. for at least one period) within the tripping
the differential protection in the very high zone near to the fault characteristic. The add-on
short-circuit current range in the event of exter- restraint operates separately for each phase but
nal faults; it is based on the setting value can be extended to blocking all phases depending
EXF-Restraint (address 1261) and has the slope 1 on the used vector group (crossblock function).
(address 1241). The recommended setting value for the crossblock
function is 15 periods (preset).
n The slope 2 together with base point 2 lead to
higher stabilization in the higher current range Notes on setting the inrush blocking
at which current transformer saturation can oc- An inrush current with a high proportion of 2nd
cur. The slope of this characteristic section is set harmonics is generated when switching the trans-
to 50 %. former on, which can lead to false tripping of the
n The IDiff >> stage works without restraint (stabi- differential protection. The preset value for the in-
lization) and is designed for high internal fault rush restraint with 2nd harmonics of 15 % can be
currents on the primary side of the transformer accepted unchanged. A lower value can be set for
with a high degree of saturation. It should be set greater restraint in exceptional cases under un-
to at least 20 % above the max. through flowing favorable energizing conditions resulting from the
fault current or the max. inrush currents, respec- design of the transformer.
tively.
The inrush restraint can be extended by the cross-
Notes on add-on restraint block function. This means, that all three phases
of the IDiff > stage are blocked when the harmonic
In systems with very high traversing currents, a component is exceeded in only one phase. A set-
dynamic add-on restraint becomes effective for ting value of 3 periods, effective for the time of
external faults. Presetting 4.0 can, as a rule, be ta- mutual blocking after exceeding the differential
ken over without change. The value is referred to current threshold, is recommended (preset).
the rated current of the protected object.
Note that the restraint current is the arithmetic
sum of the currents flowing into the protected ob-
ject, i.e. is double the traversing current. The

Notes on setting the overexcitation blocking With parameterization in secondary values the
currents in amperes are converted to the secon-
Stationary overexcitation in transformers is char-
dary side of the current transformers.
acterized by odd harmonics. The third or fifth
harmonic is suitable for restraint. Since the third Secondary setting value:
harmonic is often eliminated in transformers (e.g.
4375 A
in a delta winding), the 5th harmonic is mostly I >> = ⋅ A = 8.75 A
used. The proportion of 5th harmonics which leads 500 A
to blocking of the differential protection is set at i.e. at short-circuit currents above 4374 A (prima-
30 % (preset). It is not usually necessary to set the ry) or 8.8 A (secondary), there is definitely a fault
crossblock function in this case. in the transformer area, which can be eliminated
immediately by the overcurrent time protection.
5.3 Backup protection functions Increased inrush currents must be considered as
5.3.1 Overcurrent time protection well. The inrush restraint does not affect the stage
The definite-time overcurrent-time protection of I>>.
the 7UT612/613 serves as backup for the short-
circuit protection of the downstream power sys- Stage I> represents the backup protection for the
tem sections when faults cannot be cleared in time downstream busbar. It is set higher than the sum
there, meaning that the protected object is in dan- of the rated outgoing currents. Pickup by over-
ger. load must be ruled out because the relay operates
The overcurrent-time protection can be assigned with correspondingly short command times as
to one of the three voltage sides of the transform- short-circuit protection in this mode and not as
er. Correct allocation between the measuring in- overload protection. This value must be converted
puts of the relay and the measuring locations (cur- to the higher-voltage side of the transformer. The
rent transformer sets) of the power plant must delay time depends on the grading time in the
also be observed. The stage I>> together with sta- outgoing lines. It should be set e.g. 200 ms more
ge I> or stage IP produces a two-stage characteris- than the greatest grading time on the LV side.
tic. If the overcurrent-time protection acts on the Moreover, the inrush restraint for the I> stage
feed side of the transformer, stage I>> is set so must be parameterized effectively in this case, so
that it picks up for short-circuits extending into that false pickup of the I> stage (resulting from
the protected object, but not for a short-circuit the inrush of the transformer) is prevented.
current flowing through it. 5.3.2 Overload protection
Calculation example: The thermal overload protection prevents over-
loading of the transformer to be protected. Two
Autotransformer YNyn0 methods of overload detection are possible in the
50 MVA 7UT6:
66 kV/33 kV
uSC = 12 % n Overload protection with thermal replica accor-
Current transformer 500 A/1 A on the 66 kV side ding to IEC 60255-8,
The overcurrent-time protection acts on the n Hot-spot calculation with determining of the
66 kV side (= feed side). relative ageing rate according to IEC 60354

The maximum possible three-phase short-circuit One of these two methods can be selected. The
current on the 33 kV side with rigid voltage on the first is notable for easy handling and a low num-
66 kV side would be: ber of setting values; the second requires some
knowledge of the protected object, its ambient
1 1 S N Transfo
I 3 pole max = ⋅ I N Transfo = ⋅ context and its cooling, and needs the input of the
uSC Transfo uSC Transfo 3U N coolant temperature via a connected thermobox.
The second method is used when the transformer
is operated at the limit of its performance and the
1 35 MVA
= ⋅ = 3645 A relative ageing rate is to be monitored by the
012
. 3 ⋅ 66 kV hot-spot calculation.
With a safety factor of 20 % this gives the primary Overload protection with thermal replica (to act on
setting: the HV side) is chosen for this application. Since
the cause of the overload is normally outside the
I>> = 1.2 x 3645 A = 4374 A
protected object, the overload current is a traver-
sing current. The relay calculates the temperature
rise according to a thermal single-body model by
means of the thermal differential equation

2 5.3.3 Overexcitation protection

dΘ 1 1  I 
+ ⋅Θ = ⋅  The overexcitation protection serves to detect in-
dt τ th τ th  k ⋅ I N Obj  creased induction in generators and transformers,
especially in power station unit transformers. An
The protection function therefore represents a increase in the induction above the rated value
thermal replica of the object to be protected quickly leads to saturation of the iron core and
(overload protection with memory function). high eddy current losses which in turn lead to im-
Both the history of an overload and the heat trans- permissible heating up of the iron.
mitted to the ambient area are taken into account. Use of the overexcitation protection assumes that
Pickup of the overload protection is output as a measuring voltages are connected to the relay. The
message. overexcitation protection measures the voltage/
frequency quotient U/f, which is proportional to
Notes on the setting
the induction B at given dimensions of the iron
In transformers, the rated current of the winding core. If the quotient U/f is set in relation to voltage
to be protected, which the relay calculates from and frequency under rated conditions of the pro-
the set rated apparent power and the rated voltage, tected object UN Obj/fN, a direct measure is obtain-
is significant. The rated current of the side of the ed of the induction related to the induction under
main protected object assigned to the overload rated conditions B/BN OBj. All constant
protection is used as the basic current for detec- variables cancel each other:
ting the overload. The setting factor k is deter-
mined by the ratio of the thermally permissible U
continuous current to this rated current: B U N Obj U/f
= =
I max BN Obj f U N Obj / f N
k= fN
I N Obj
The relative relation makes all conversions unnec-
The permissible continuous current is at the same
essary. All values can be specified directly related
time the current at which the e-function of the
to the permissible induction. The rated variables
overtemperature has its asymptote. The presetting
of the protected object have already been entered
of 1.15 can be accepted for the HV winding.
in the 7UT613 relay with the object and trans-
Time constant τ in thermal replica: former data when setting the differential protec-
The heating time constant τth for the thermal rep- tion.
lica must be specified by the transformer manu- Setting instructions
facturer. It must be ensured that the time constant
The limit value of permanently permissible induc-
is set in minutes. There are often other specifica-
tion in relation to the rated induction (B/BN)
tions from which the time constant can be deter-
specified by the protected object manufacturer
mined:
forms the basis for setting the limit value. This
Example: value is at the same time a warning stage and the
t6 time: This is the time in seconds for which minimum value for the thermal characteristic Fig. 13
6 times the rated current of the transformer wind- (see Fig. 13) Tripping characteristic of the
ing may flow. overexcitation protection
τ th
= 0.6 ⋅ t 6
min
If the transformer winding has a t6 time of 12 s
τ th
= 0.6 ⋅12 s = 7.2
min
the time constant τ must be set to 7.2 min.

An alarm is output after of the relevant set delay Service interface

time (about 10 s) of the overexcitation stage U/f The 7UT613 has a separate service interface which
has expired. Major overexcitation endangers the can be read out by telecommunication via a mo-
protected object already after a short time. The dem. The user is informed in the office quickly
high-set, fast tripping stage U/f>> is therefore set and in detail about the transformer fault. The data
to a maximum of 1 s. are then analyzed in the office by DIGSI. If this re-
The thermal characteristic should simulate the mote fault clearance is insufficient, the fault data
heating, i.e. temperature rise, of the iron core re- provide hints for an efficient service mission.
sulting from overexcitation. The heating charac-
teristic is approximated by entering 8 delay times n 7. Summary
for 8 given induction values B/BNObj (referred to in Optimum protection of the transformer with
simplified form as U/f). Intermediate values are SIPROTEC relays means security of investment for
obtained by linear interpolation. If no data are valuable operating equipment and therefore makes
available from the protected object manufacturer, a contribution to maximum supply security.
the preset standard characteristic is used. From a technical point of view, the 7UT612 or
7UT613 relay offers extensive short-circuit protec-
n 6. Further functions tion for both the main and the backup protection
6.1 Integration in substation control and of transformers in one single relay. Further
protection SIPROTEC relays supplement the main protecti-
The protection can be connected to a substation on and enhance the reliability of the protection
control system via the system interface and oper- scheme by way of hardware redundancy.
ated in parallel by PC via the service interface to a
star coupler for separate remote communication. Extensive measuring functions allow trouble-free
connection of the relay without the need for addi-
tional equipment, and enable monitoring of the
transformer in operation in terms of its electrical
and thermal parameters. The relay’s presettings
are chosen so that the user need only parameterize
the known data of the main and primary trans-
formers. Many of the default settings can simply
be accepted as they are, and therefore make para-
meterization and setting easier.
Fig. 14 Integration in substation control and protection
Via the interface,

n Messages
n Alarms
n Measured values
are transmitted from the transformer differential

protection to the substation control system. Mes-
sages are available for every one of the activated
protection functions, which can be either trans-
mitted to the substation control system in the
course of plant equipment parameterization, or
configured to the LEDs or message contacts in the
protection relay. Configuration is made clear and
easy by the DIGSI matrix.

Motor Protection
Protection of a
Motor up to 200 kW
n 1. Introduction
Drive motors often play a decisive role in the
functioning of a production process. Motor dam-
age and breakdowns not infrequently lead also to
consequential damage and production shutdowns,
the cost of which significantly exceeds the cost of
repairing the motor. Optimum design of the mo-
tor protection ensures that damage following
thermal overload is prevented, meaning that there
is no reduction to the normal service life. Second-
ary faults are minimized in the event of short-
circuits, earth faults and winding faults.
The spectrum extends from small low-voltage
motors with an output of a few kW to high-
LSP2731.tif
voltage motors with outputs measured in MW.
Protection system design must be based on the
rating of the motor, the importance of the drive Fig. 1 SIPROTEC 7SJ602 multifunction protection
for the technological process, the operating condi-
tions and the requirements of the motor manufac- Such overloading cannot and should not be de-
turer. tected by short-circuit protection since any poten-
The setting of a SIPROTEC protection relay for tial delay must be very short on these occasions.
motor protection is described below taking a Overload protection prevents thermal overloading
high-voltage motor (10 kV) as the example. of the motor to be protected. The 7SJ602 relay de-
tects stresses either before overload occurs (over-
n 2. The tasks of motor protection load protection with complete memory = thermal
Motors have some striking features in their oper- replica) or only after exceeding a preset start-up
ating conditions. These are important for under- current (overload protection without memory
standing the various possible causes of failure and function).
must be taken into account when designing pro- n Overload protection without memory
tection systems. If overload protection without memory is cho-
2.1 Protection of the stator against thermal sen, the tripping time is calculated according to
overload a simple formula. Pre-stressing is not taken into
The power drawn by the motor from the supply account because currents are only recorded if
system during operation is supplied to the shaft as they are greater than 1.1 times the set value.
mechanical power for the production machine. 35
t= ⋅ t 6IB for I > 1.1 IB
The power lost to the winding during this energy (I / I B )2 − 1
conversion is the decisive factor for the arising t Tripping time
motor temperatures. The loss of heat is propor- I Overload current
tional to the square of the current. The motor IB Set threshold
heating time characteristic is determined by its t6IB Set time factor
heat storage capability and heat transfer proper- (t6-time = tripping time when applying 6 times
the actual set value IB)
ties, and characterized by the thermal time con-
stant [τ]. Electrical machines are at particular risk
from long-term overload. Thermal overloading of
the motor leads to damage to the insulation and
therefore to secondary faults or to a reduction in
the total service life of the motor.

Motor Protection
n Overload protection with memory permitted start-ups from cold (nc) and operat-
The relay calculates the temperature rise in ac- ing temperature (nW) conditions. The relay cal-
cordance with a thermal homogenous-body culates from this the value of the thermal rotor
model and a thermal differential equation. In replica and gives a blocking command until the
this way the previous load, with all load cycles, thermal replica of the rotor reaches a value be-
can be recorded and evaluated correctly by the low the restart limit and therefore permits a new
relay. Such a thermal replica can be optimally start-up. As long as a blocking command pre-
adapted to the overload capacity of the pro- vails, switching on by the relay’s integrated
tected equipment. switch control is prevented. In this case, it is not
necessary to allocate the restart inhibit’s block-
2.2 Protection of the rotor from thermal overload ing command to a command relay or an exter-
Among the many causes of excessive temperatures nal link with the switch control. If however the
caused by currents in motors is an unacceptably motor can be switched on from another posi-
long start-up time or, in limit cases, blocking of tion, an output relay must be allocated to the
the rotor. Such conditions are caused by an exces- blocking command and its contact looped into
sive mechanical load torque, such as can occur in the starting circuit.
overfilled mills and breakers or overloaded centri-
fuges, etc. 2.3 Negative-sequence protection
In protection of the motor, negative sequence
n Start-up time monitoring
(unbalanced load) protection assumes particular
The protection relay has start-time monitoring,
importance. Unbalanced loads produce a reverse
which represents a meaningful addition to over-
field in motors which drives the rotor at twice the
load protection for electrical machines. The trip
frequency. Eddy currents are induced on the sur-
time depends on the current. This enables even
face of the rotor, leading to local temperature rises
extended start-up times to be correctly evalu-
in the rotor. If the motor is protected by fuses, a
ated when the start-up current is reduced be-
phase voltage failure is a frequent fault in practice.
cause of voltage sags when the engine is started.
During this breakdown, the line-to-line voltage is
The start-up time monitoring begins when a set
fed to the stator winding by the remaining work-
current level is exceeded. The trip time depends
ing phases. Depending on the load, a more or less
on the actual measured start-up current. If the
circular rotating field is maintained by the motor,
permissible locked rotor time is shorter than the
so that it can develop sufficient torque with in-
start-up time, the rotational speed (motor sta-
creased current input. There is also the risk of
tionary or rotating) must also be requested via a
thermal overload if the system voltage is unbal-
binary input.
anced. Even small voltage unbalances can lead to
n Restart inhibit large negative sequence currents because of the
Restart inhibit prevents the motor restarting if, small negative sequence reactance.
during this start-up, the permissible rotor heat-
ing is expected to be exceeded.
The rotor temperature of a motor generally lies
well below its permitted temperature limit both
during normal operation and with increased
load currents. On the other hand, during
start-up, with the associated high start-up cur-
rents, the rotor is at a higher risk of thermal
damage than the stator because of its smaller
thermal time constant. The motor must be pre-
vented from switching on if the permissible ro-
tor heating is expected to be exceeded during
this start-up. This is the task of the restart in-
hibit. Because the rotor current is not directly
measurable, stator currents must be relied upon
from which the rotor temperature is indirectly
calculated. It is therefore assumed that the ther-
mal limit values for the rotor winding in the
data provided by the motor manufacturer for
the rated start-up current equal the maximum
permitted start-up time and the number of the

Motor Protection
The 7SJ602’s negative-sequence protection filters

the fundamental out of the fed phase currents and
breaks it down into balanced components (nega-
tive phase-sequence system I2 and positive phase-
sequence system I1). The “negative phase-sequence
system/rated current (I2/IN)” relationship is as -
sessed to detect the unbalanced load. The nega-
tive-sequence protection is set up in two stages.
After reaching a first, adjustable threshold I2> a
time stage TI2>is started; after reaching a second,
adjustable threshold I2>> the time stage TI2>> is
started. After one of the operating times has
passed a tripping command is set up. The thresh-
old comparison can only be made if the largest of
the three phase currents is at least 10 % of the
rated current.
Fig. 3 Connection of 4 CTs with measurement of the earth current
7SJ6022../7SJ6026.. for resonant-earthed, isolated

or high-resistance earthed power systems
The relay has two phase current transformers, a
sensitive earth current input and a voltage input,
Fig. 2 Characteristics of the negative sequence protec-
tion e.g. to record Uen Voltage.
2.4 Earth-fault protection

When designing earth-fault protection it is impor-
tant to know how the star point of the power sup-
ply system is connected. This must also be taken
into account when selecting protection relay hard-
ware. Two versions of the 7SJ602 are available,
differing in the design of the transformer inputs.
7SJ6021../7SJ6025.. for low-resistance earthed
power systems
The relay has a fourth input transformer with
“normal” sensitivity for recording the earth cur -
rent. This can be connected to the star point lead
wire of the current transformer unit or to a sepa-
rate core-balance current transformer.
Fig. 4 Connection of 3 CTs and 1 VT with measurement of the earth current and one
phase voltage

Motor Protection
Earth-fault protection detects earth faults in the dimensioning the setting in order to be independ-
stator winding of three-phase machines. Because ent of the circuit state in the power supply system.
motors are usually connected directly by a busbar With machines in compensated power supply
to a power supply system (directly connected to a systems a load device and measurement of the dis-
busbar) it is important to recognize whether the placement voltage are always recommended so
earth fault is in the machine feeder or on another that a safe earth-fault decision can be made.
feeder of the busbar.
2.5 Short-circuit protection
With earthed systems, this can usually be clearly
recognized from the magnitude of the earth cur- The task of the short-circuit protection when a
rent. When a fault occurs in the machine, the full short-circuit occurs is both to prevent increased
earth-fault current driven by the power supply damage to the motor (destruction of the iron
system flows via the protection measuring point. core, etc.) by quickly switching off the motor and
The machine must be isolated from the power to minimize the effect on the power supply system
supply system as quickly as possible to prevent with its connected loads (voltage unbalance, volt-
more damage. When there is a power system age sags, etc.).
earth-fault the recorded earth current is essentially The overcurrent-time protection in the 7SJ602
determined by the machine capabilities and can take the form both of definite-time overcur-
therefore considerably smaller. There must be no rent-time protection and of inverse-time over-
tripping. current-time protection. For the latter, a range of
In compensated, isolated and low-resistance characteristics defined in IEC 60255---3 or in
earthed systems, a design with sensitive earth-cur- ANSI standards is available. A high-current stage
rent input and sensitive earth-fault detection I>>, which always works with definite tripping
should be chosen. The high-resistance earth-fault time, can be superimposed on the selected over-
detection then replaces the earth-current stage of current characteristics. An instantaneous tripping
the overcurrent-time protection. Because of its stage I>>> can also be superimposed on the phase
high sensitivity it is not suitable for detecting earth branches. In this way the tripping characteristics
faults with large earth currents (more than around can be optimally adapted to the motor's start-up
1.6 ⋅ IN on the terminals for sensitive earth-cur- characteristics.
rent connection). In order to be able to switch off during high cur-
Overcurrent-time protection for earth currents rent faults in the machine, the 7SJ602 has a special
must be used here. instantaneous tripping stage. The I>>> stage must
Should the magnitude of the earth current be suf- be set safely above the motor’s inrush current, so
ficient to determine the earth fault, no voltage in- that switching on the motor does not lead to trip-
put is needed. The 7SJ602 has a two-stage current/ ping. Experience has shown that the inrush cur-
time characteristic which works with earth-currents can be around 1.5 to 1.6 times the start-up
rent values. They are appropriate where the mag- current.
nitude of the earth current enables the location of
the earth fault to be defined. This can, for exam- The I>> stage should be set above the motor start-
ple, happen with machines on low-resistance up current to prevent tripping. With the time de-
earthed systems (with earth-current limiting). lay TI>> the period of the inrush current must be
taken into account. Because the inrush current
With machines directly connected to a busbar to lasts only a few ms, the TI>> can be selected at
isolated power systems, it is essential that the ca- around 50 ms.
pacity of the upstream power system delivers a
sufficiently large earth current but that the earth
current at the relay location is comparably small
in the case of earth faults on the power system
side. The magnitude of the earth current is used to
reach a decision on the position of the fault
location.
If this is not the case an additional earth-current
production device must be installed on the
busbar. This produces a defined earth current
during an earth fault. The connected displacement
voltage is then used to make a direction decision.
Should a load device (earth-current production
device) be installed, it should only be used when

Motor Protection
In the overcurrent-time protection function an value applies for all three phases. If the set current
inverse-time characteristic must be chosen since value in one phase is exceeded, the circuit-breaker
this can be better adapted to the motor’s opera- is considered closed. In machines the value se-
tional performance. lected must be smaller than the machine’s mini-
mum no-load current.
The inverse-time short-circuit protection Ip> pro-
tects the motor from short-circuits during opera- Motor data is generally related to the rated motor
tion in transient condition (after ramping up). current. A matching factor must be communi-
The higher the short-circuit current the quicker cated in the system data to the 7SJ602 so that the
the tripping. The extreme inverse characteristic settings for motor protection functions can be
must be selected as tripping characteristic. provided directly as a reference quantity.
Example:
Current transformer 100 A / 1 A
Rated motor current INM = 74 A
n Ratio of rated motor current to rated trans-
former current
Im = INM/INTRANSF = 0.74 [from transformer
data]
The motor’s start-up current is likewise preset
in the 7SJ602's system data.
The start-up current is specified as a value re-
lated to the rated motor current (INM).
It depends on the size and nature of the motor
Ip> Short-circuit protection, inverse-time stage and in a normal load-free start-up is approxi-
I>> Short-circuit protection, definite-time stage mately 5 ⋅ INM.
I>>> High-set instantaneous tripping stage,
definite-time stage n Motor start-up current referred to rated motor
Fig. 5 Current characteristic of motor start-up
current
Ia = 5 [from motor data sheet]
In the 7SJ602, the motor’s start-up time is pre-
n 3. Adjustments set in the system data. After this time the start-
Calculation examples are oriented towards the fol- up current must be safely undershot.
lowing motor data: n Motor start-up time
Motor/system data tSTART-UP = 4.3 [from motor data sheet]
Current transformer phase
INPRIM/INSEC 100 A / 1 A 3.1 Overload protection
Current transformer earth For overload protection the load must be taken
(60/1) IEE/IPH 0.6 (core- into account before the overload occurs; i.e. the
balance CT) overload function must be used with full memory.
Voltage transformer 10 kV / 100 V
The relay calculates the temperature rise in accor-
Motor rated current INM 74 A
dance with a thermal homogenous-body model
Max. permissible unbalanced load 10 %
and a thermal differential equation:
Permissible unbalanced load period 15 s
Permissible continuous thermal dΘ 1 1 2
+ ⋅Θ = ⋅I
current IMax 1.1 ⋅ INM dt τ τη τ th
Thermal stator time constant τth 40 min
Standstill transient factor kτ 5 Θ Present temperature rise referred to the final
temperature with maximum permissible line
Start-up current IA 5 ⋅ INM current k · IN
τth Thermal time constant for heating of the object
Data of the system and equipment to be protected to be protected
is input. Some data which particularly involves I Present effective current referred to the maximum
motor protection functions is worthy of mention permissible current Imax =k · IN
here.
For some protection functions it is important to
recognize whether the circuit-breaker is closed or
open. As a criterion for this overshooting or un-
dershooting a current threshold is applied. The set

Motor Protection
The following parameters must be set: The following parameters must be set:
n Set value of the k factor = Imax/INTransf n Start-up current threshold Ia> for start-up time
Imax = maximum permissible continuous monitoring, referred to rated motor current INM
thermal current = 1.1 ⋅ INM = 81.4 A with Ia = 5 ⋅ INM motor data entered with system
k = 0.82 data
n Set value of the thermal time constant τth in Ia> = 0.5 Ia = 2.5 INM
minutes Should the start-up time exceed the tripping time
th = 40 min [from motor data sheet] of the overcurrent time protection, said protec-
For motors the t6 time, i.e. the permissible time tion is blocked during start-up after 70 ms.
for the six-fold permissible continuous current, is n Blocking the I> / Ip stages during start-up
often specified instead of the time constant. NO
As a result the τth is calculated as follows:
If the permissible locked rotor time is less than the
t6 start-up time, the rotational speed (engine stands
t
Set value τ th[min] = s ⋅ 36 = 0.6 ⋅ 6 or rotates) must be additionally requested via a bi-
60 s nary input.
n Transient factor kτ between time constant
(during standstill) and running of the motor 3.3 Restart inhibit
k = 5 according to motor data Rotor temperature simulation plays a decisive role
n Alarm temperature rise as a percentage of the in the restart limit. The parameters required for
operating temperature rise ΘALARM/ΘTRIP this such as start-up current, rated motor current
ALARM = 90 % [preset]
and maximum permissible start-up time are con-
figured with the system data.
The 7SJ602 also provides the option to connect an
external thermobox to the relay. This affords an The following parameters must also be set in
opportunity to connect the coolant temperature addition during restart inhibit:
or ambient temperature of the protected object n Temperature equalization time of the rotor
into the relay using the serial interface and include As the thermal time constant of the rotor is
it in the overload calculation. considerably smaller than that for the stator, a
value of 1 min. at most (preset) is practicable
3.2 Start-up time monitoring for the temperature equalization time of the ro-
The start-up time monitoring interprets over- tor.
shooting the current value Ia> as a motor start-up. tEQUAL = 1min [empirical value]
Consequently this value must be chosen so that it n Number of maximum permissible warm
is safely exceeded during motor start-up in all load start-ups
and voltage conditions by the actual start-up cur- nW = 2 [empirical value]
rent but is not reached during permissible, short-
term overload. It must also be configured above If no specifications are available from the motor
the maximum load current. The set value is re- data sheet, empirical value 2 is set.
lated to the rated motor current. Half the value of n Difference between the number of maximum
the rated start-up current is customary. If the permissible cold start-ups and the maximum
start-up current is 5 ⋅ rated motor current (INM), number of permissible warm start-ups
Ia> is set at 2.5 ⋅ rated motor current. nc – n W= 1 [empirical value]
The tripping time is calculated quadratically
according to the magnitude of the current: If no specifications are available from the mo-
tor data sheet, empirical value 1 is set.
2
t TRIP = t START -UP ⋅  a  with I > Ia>

I n Factor for the thermal cooling-down time of the
 I  rotor when the machine is at standstill.
tTRIP Actual tripping time for flowing current I The reduced cooling (when the motor is at
tSTART-UP Max. start-up time standstill) in motors with self-ventilation is
I Actual flowing current (measured quantity) taken into account by the factor kτSTI (related to
Ia Rated motor start-up current the time constant during no-load operation).
Undershooting the current threshold set in the
system data as LS I> is considered as criterion
for the motor’s standstill.

Motor Protection
1 is set in forced-ventilated motors. 3.5 Earth-fault protection

kτSTI = 5 [empirical value] Earth-fault protection detects earth faults in the
If no specifications are available from the motor stator winding of three-phase machines. The de-
data sheet, empirical value 5 is set. sign of the earth-fault protection depends on how
the star point of the power supply system is con-
n Factor for the thermal cooling-down time of the nected. The star point of the motor must always
rotor when the machine is running. be set up isolated as follows. The 7SJ602 is suitable
This factor takes into account the different both for power systems with earthed star point
cooling-down of a loaded, running motor com- and for power systems with isolated, compensated
pared to that of a motor which is switched off. It or low-resistance earthed star point. Two hard-
is effective as soon as the current exceeds LS I> ware variants cover all these system configura-
(set in system data). tions. These must be taken into account when
With kτoperation = 1 heating and cooling-down ordering.
time constants are equal under normal operat-
ing conditions. The following depicts the earth-fault protection
kτBET = 2 [empirical value] setting for a motor feeder in a 10 kV isolated sys-
If no specifications are available from the motor tem.
data sheet, empirical value 2 is set. As this is an isolated system the variant must be
n Minimum lock-out time chosen with sensitive earth-fault detection.
The minimum lock-out time tLOCK relates to the
motor manufacturer's specifications or operat-
ing conditions. It must be greater than the time
of temperature equalization tEQUAL .
tLOCK = 6 min [assumption]
3.4 Negative-sequence protection

In electrical machines, negative-sequence (unbal-
anced load) protection is of particular impor-
tance.
The definite-time characteristic is set up in two
stages. When a first, adjustable I2> threshold is
reached, a pick-up signal is given and a time stage
T I2> started. When a second stage I2>> is
reached, another signal is transmitted and the
time stage T I2>> started. After one of the delays
has elapsed, a tripping command is given. The
I2>> stage is well suited to faults in the secondary 2
10 kV cable 3 x 1 x 170 mm
transformer circuit with lower sensitivity and very ICE = 1.8 A / km
short tripping time, for example. l = 0.1 km
The preset values for pick-up and time delay are

mostly sufficient. If the machine manufacturer has
stated values concerning continuous permissible
unbalanced load and the duration of loadability Fig. 6 Application example motor protection
as a function of the magnitude of the unbalanced
load, these should have priority.
The percentage values are related to the trans-
former’s rated current.
n Pick-up value of stage I2> (related to rated
transformer current IN)
I2> = 10 % [from motor data sheet]
n Tripping delay of stage I2>
TI2 = 15 s [from motor data sheet]
n Pick-up value of stage I2>>
I2>> = 50 % [preset]
n Tripping delay of stage I2>>
TI2>> = 1 s [preset]

Motor Protection
In our example it is assumed that the supplying If the earth current is insufficient for fault loca-
10 kV power system has a corresponding size and, tion, an earth-fault direction determination is
during an earth fault, a capacitive earth-fault cur- configured. In this case a voltage input (Uen) is
rent ICE of approximately 20 A flows to the fault obligatory.
location. Information about the magnitude of the
With sensitive earth-fault direction determination
capacitive earth-fault current must be requested
it is not the magnitude of the current that counts
from the power system operator. Of course, the
but rather the component of the current vertical
motor feeder also delivers an earth-fault current.
to a settable directional characteristic (symmetry
The motor feeder must be connected via a 100 m
axis). A precondition for direction determination
long 10 kV cable. The motor feeder earth-fault
is exceeding of the displacement voltage stage UE
current is calculated as follows:
and a likewise parameterizable current compo-
ICEcable = I'CE ⋅ l nent that determines the direction (active [cos ϕ]
or reactive component [sin ϕ]).
I'CE = 1.8 A/km (from the cable data sheet)
l = 0.1 km In electrical machines directly connected to
busbar on the isolated power system, cos ϕ and a
ICEcable ≈ 0.20A
correction angle of around +45° can be set for the
The following earth-fault current distribution is measuring mode, because the earth-fault current
the result. often consists of a superimposition of the capaci-
tive earth-fault current from the power system
and the resistive current of a load resistor.
3.6 Short-circuit protection

Short-circuit protection is the main protection
function of the 7SJ602. It has in all three stages for
phase currents, and two for the earth current. The
overcurrent stage (I>) can if need be work with
definite or inverse command time. High-current
tripping (I>>) and instantaneous tripping (I>>>)
always work with a definite command time.
For a motor’s short-circuit protection, it must be
ensured that the set value I>>> is lower than the
smallest (2-pole) short-circuit current and greater
Fig. 7 Earth-fault current distribution depending on the than the highest start-up current. Because the
fault location
maximum inrush current is usually below
1.6 × the rated start-up current, even under unfa -
From the current distribution it is clear, that the vorable circumstances, the following setting con-
fault can be unambiguously located from the ditions apply for the short-circuit stage I>>>.
magnitude of the earth-fault current.
1.6 x IStart-up < I>>> < Ik2pole
In the event of an earth fault on the busbar, only
the 7SJ602 in the incoming feeder bay may pick As safety distance, the setting value should be se-
up. If the fault location is in the motor feeder, the lected approx. 30 % above the expected start-up
7SJ602 must pick up both in the incoming feeder current.
bay and in the motor outgoing feeder. In this ex- I>>> = 2.0 · IA = 2.0 · 5 · INM = 2.0 · 5 · 74/100 · IN
ample, therefore, a statement of the situation can ≈ 7.4 ⋅ IN
only be made via the magnitude of the earth
current. n Pickup value of the instantaneous release stage
I>>> = 7.4 ⋅ IN
The following setting recommendation for the
motor outgoing feeder has been produced as a
pickup threshold: For safety reasons ICEtotal must
not be assumed since, for example, this may be re-
duced as a result of power system disconnections.
As a base IEE> ≈ 0.5 ⋅ ICE can be chosen.
Should an earth fault not lead to tripping, the IEE>
function can also be adjusted only to signals. If no
second pickup threshold is needed, IEE>> can be
deactivated.

Motor Protection
The setting of I>> is aligned to the motor start-up n 4. Summary

current. A safety factor of around 1.5 must be set In this example for setting, it is evident that a
in order that correct start-up does not lead to trip- SIPROTEC relay can provide comprehensive pro-
ping. tection of a motor, and that thereby (in addition
1.5 ⋅ IStart-up < (I>>) < (I>>>) to actual protection of the equipment) protection
of the remaining power supply system is also
I>> should be set above the motor start-up cur- available.
rent so that it is not tripped by it.
From a protection point of view, the relay offers
I>> = 1.5 · IA = 1.5 · 5 · INM = 1.5 · 5 · 74/100 · IN extensive protection functions for low power class
≈ 5.5 IN motors in addition to short-circuit protection. Be-
n Pick-up value of the high-current stage cause all protective functions are available in one
I>> = 5.5 ⋅ IN relay, connection and testing costs are minimized.
The time delay for the high-current stage should The relay presettings are selected so that the user
be delayed until the maximum inrush current has can apply many parameters, even if they are not
safely decayed. The values in the motor data sheet well known. Motor sheet data is mainly used to
must always have priority over the assumptions select the required values, thus making setting
and empirical values used in these applications. easier.
n Tripping delay for the high-current stage
TI>> = 50 ms
In the overcurrent protection function an inverse-
time characteristic must be chosen since this can
be better adapted to the motor’s operational
performance.
The inverse-time short-circuit protection Ip pro-
tects the motor from short-circuits during normal
operation (after start-up). The higher the short-
circuit current the quicker the tripping. The ex-
treme inverse-time characteristic must be selected
for tripping.
The maximum normal current is essential for set-
ting the overcurrent stage Ip. Pickup by overload
must be ruled out, because with this mode the
relay works with correspondingly short command
times as a short-circuit protection, not as an over-
load protection.
Ip = 1.5 ⋅ INM/1.1 = 1.5/1.1 ⋅ 74/100 ⋅ IN ≈ 1.0 IN
n Time multiplier for phase currents
Tp = 1.5 s
n Set value of the overcurrent stage Ip for the
phase currents
Ip = 1.0 IN

Motor Protection

Protection of a Medium-
Sized Generator up to 5 MW
n 1. Introduction
Small-scale stations are making a significant con-
tribution to power generation. Hydropower plants
currently still account for the largest share of sys-
tem infeed. The most significant increase in the
number of generating plants has been in wind
energy.
Electrical protection is essential for the reliable
LSP2171-afpen.eps
operation of such equipment.
The scope of protection must be in proportion to
the overall costs and importance of the plant. The
scope and choice of protection functions are influ-
enced by plant type, generator design and addi-
tional equipment, output level and power system
connection. The following table gives an overview
of the protection functions used depending on Fig. 1 SIPROTEC 7UM generator, motor and transformer
protection
generator output.
For hydropower generators For diesel generators and turbogenerators

Up to 300 to 700 to > 1500 up to 300 to 700 to > 1500
300 700 1500 300 700 1500
kVA kVA kVA kVA kVA kVA kVA kVA
Thermal and short-time delayed x – – – x – – –
trip and shunt release for U~
on generator circuit-breaker
Only shunt releases for U~ on – x x x – x x x
generator circuit-breaker
Rise-in-voltage protection x x x x – – x x
Reverse-power protection – – – – x x x x
Overcurrent-time protection – x x x – x x x
Differential protection – – – x – – – x
Rotor earth-fault protection – – – x – – – x
Is DC auxiliary voltage for – x x x – x x x
protection required?
Table 1 Protection functions for small-scale power stations

In small-scale power stations, the basic circuits for
busbar and unit connection (as shown in Fig. 1)
can be assumed.
Unit connection Busbar connection
Fig. 2 Plant basic circuit
Fault type Cause Protection function Remarks

In unit connection, the generator is linked to the
Overload Sab>Sproduced Thermal overload Evaluation of current higher voltage level busbar via a transformer. In
2
Controller error protection (I t) r.m.s. value with pre-
Maloperation vious load recording
the case of several parallel units, the generators are
electrically isolated by the transformers.
Short-circuit Deterioration of insula- Overcurrent-time Time delay must be
(2 or tion protection (I>) coordinated with In busbar connection, several generators feed
3 phase) Winding displacement Differential protection system protection onto a common busbar. Subsequently, the next
Overvoltages (∆I) higher voltage level is fed via a transformer. The
Manufacturing defects generators are galvanically connected.
Earth fault Same cause as for short- Stator earth-fault pro- The protected zone
(Stator) circuit tection U0> in unit (approximately
Owing to the low overall plant costs, the busbar
connection 80 %) is determined connection is frequently chosen for small-scale
Earth-fault direction by plant conditions power stations. This application is therefore con-
in busbar connection (see discussion in text)
(/UE, IE)
sidered in greater detail in the following.
Earth fault Deterioration of insula- Rotor earth-fault pro- Used as from 5 MW, Table 2 shows the protection functions suitable
(Rotor) tion tection with system if sliprings available; for small-scale power stations in accordance with
Winding displacement frequency signal cou- below 5 MW optional today’s state-of-the-art. Fault type, cause and the
pling in rotor circuit
Brush abrasion on the protection function to be deployed are indicated,
slipring surface
together with general notes on particular features
Material fatigue
of the protection function.
Reverse- Drive failure Reverse-power Only necessary for
power Shutdown protection (-P) steam and diesel drive
systems
Speed Leaking stream valves Frequency protection As from 5 MW f> and
irregularities Sudden changes in active (f> or f<) f<
power Below 5 MW so far
Overload only f>; f< is likewise
recommended if avail-
able
Overvoltage Controller error or Overvoltage protec- Evaluation of phase-
manual maloperation tion (U>) to-phase voltage
Unpermissible Fault in exciter circuit Underexcitation Used as from 5 MW
under- Operation in underex- protection (e.g. -Q, or Below 5 MW so far
excitation cited state (high reactive Z) not usual;
power demand in recommended if func-
system) tion available
Maloperation,
controller error
Asymmetric Unequal loading of con- Negative-sequence Used from 5 MW;
load ductor (or load unbalance) Below 5 MW so far
protection (I2>) not usual; recom-
mended if possible, Table 2 Fault type, protection functions

n 3. Applications Protection functions ANSI 7SJ60 7SJ61 7SJ62 7SJ63/64 7UM61

Table 1 shows, that even with small generators of
Rotor overload protection 49 X X X X X
< 5MW, relays must be used with a number of
protection functions. Numerical protection relays Earth-fault protection 64G X X X X
directional / non directional 50G X X X X X
are the current state-of-the-art. 67G X X X X
The SIPROTEC range provides a good choice.
Overcurrent-time protection 50 X X X X X
As shown in Table 3, 7SJ relays are well suited for 51
simple protection functions for small generators. Negative-sequence protection 46 X X X X X
1) 1) 1) 1)
Rotor earth-fault protection 64R X X X X
The decisive advantage of the 7UM6* generator
2)
protection relay is the automatic adjustment of Reverse-power protection 32 X X
the sampling frequency. To ensure that the pro- Overcurrent protection 59 X X X
tection and measurement functions deliver cor- Underexcitation protection 40 X
rect results over a wide frequency range, the actual Frequency protection 81 X X X
frequency is continuously measured and the mea-
Temperature monitoring 38 X X X X
surement processing sampling frequency continu- (by an external monitoring
ously tracked. This ensures the measuring accura- box called thermo-box)
cy in the frequency range from 11 Hz to 69 Hz. Breaker failure protection 50BF X X X X X
This relay offers a wide range of additional protec-
Programmable logic X X X X
tion functions. If differential protection is re-
quired and the appropriate transformer sets are Control functions X X X X X
available, a 7UM62 is recommended. As differen- Flexible serial interface 1 2 2 2/3 2
tial protection is applied usually for generators 1) via IEE measuring input if earth-fault direction function is not used.
above 5 MW, the example shown opposite refers 2) in 7SJ63 with CFC, in 7SJ64 with flexible functions.
to a 7UM61 relay for a 5 MW generator in busbar
Table 3 Protection relays – selection matrix
connection.
n 4. Settings
In the following sections, the individual protec-
tion and additional functions (see Table 3) are
explained. Notes on the setting values are also
given. The calculation examples are oriented to-
wards the reference plant shown in Fig. 3. For the
tripping concept, it is assumed that the protection
directly actuates the tripping (circuit-breaker,
de-excitation, turbine valve closing or diesel
cut-off).
4.1 Thermal overload protection (ANSI 49)

Overload protection prevents thermal overload of Fig. 3 Busbar connection with core-balance CT
the stator windings on the machine to be protec-
ted. The relay calculates the temperature rise in Low ambient or coolant temperatures mean that
accordance with a thermal single-body model by the generator can be loaded with more current;
means of the thermal differential equation and high temperatures signify that the loadability is
takes account of both previous overload history less.
and emission of heat into the ambient area.
Example
After an initial, adjustable threshold has been Generator and transformer with the following
reached, an alarm signal is emitted for the purpose data:
of enabling a load reduction in good time, for n Permissible continuous current
example. Imax prim = 1.15 • IN, generator
The second temperature threshold disconnects n Rated generator current IN, generator = 483 A
the machine from the system. For example,
n Current transformer 500 A/1 A
ambient or coolant temperatures can be input via
the PROFIBUS-DP interface.

Set value k-factor = 1.15 483A/500A = 1.11 4.3 Definite-time overcurrent-time protection
(I>) with undervoltage seal-in (ANSI 51V)
Note:
Taking the k factor at the usual figure of 1.1, ap- Setting example:
plying the generator rated current (with the pri- Pickup value 1.4 ⋅ INGenerator
mary transformer current matched) produces a Tripping delay 3 s
temperature rise of Θ/ ΘK = 1 / 1.12 = 0.83 of the Undervoltage seal-in 0.8 UNGenerator
tripping temperature. The alarm stage should thus Seal-in time of U < 4 s
be set between end temperature at rated current Dropout ratio 0.95
(83 % in this case) and tripping temperature
(100 %). 4.4 Earth-fault protection
In addition to short-circuit protection, which as
With an assumed load current of I = 1.5 IN (relay) described above is provided in a familiar fashion
and a preload of Ipre = 0, the following tripping via overcurrent (or differential) protection, earth-
times are derived for various ambient temperatures fault protection is of particular significance for
ΘK= 40 ºC t = 463 s small-scale machines.
ΘK= 80 ºC t = 366 s
ΘK= 0 ºC t = 637 s 4.4.1 Principle
A particular feature of electric machines with iso-
4.2 Definite-time overcurrent-time protection lated star point is that the displacement voltage
(I>, I>>) (ANSI 50/51) decreases linearly as the fault location moves in
General the direction of the generator star-point (Fig. 4).
Overcurrent-time is the form of short-circuit pro- The earth-fault current, the magnitude of which is
tection for extra-low or low voltage generators. In determined by the earth capacitances in addition
order that internal faults are always responded to, to the displacement voltage, thus also decreases.
the generator protection is connected to the In the event of faults close to the star point, the
current transformer set located in the star point displacement voltage and earth current become so
connection of the generator. In the case of genera- small that they can no longer be reliably mea-
tors whose excitation voltage is taken from the sured.
machine terminals, in the event of nearby faults A protected zone of 80 – 90 % is consequently
(i.e. in the generator or the unit transformer re- spoken of.
gion) the short-circuit current decays very quickly
Fig. 4 since there is no longer any excitation current, and In unit connection (Fig. 2a), the protected zone
Displacement voltage within a few seconds falls below the overcurrent- discussed above is additionally determined by the
as a function of the fault disturbance signal injection from the upstream
location in the stator time protection pickup value. In these cases
winding undervoltage seal-in is used. system. If an earth fault occurs in the system, a
displacement voltage is identifiable via the cou-
pling capacitance of the unit transformer. The
magnitude of the interference voltage is deter-
mined by the coupling capacitance, the genera-
tor-side earth capacitance (stator, incoming line)
and the difference between rated system voltage
and rated generator voltage.
In busbar connection, the displacement voltage
can only be used for earth-fault indication due to
the galvanical connection of the generators. The
earth-fault direction protection makes selective
tripping possible. The protected zone is deter-
mined by the earth current, which is measured by
a core-balance current transformer (60 A/1 A). As
shown in Fig. 2b the sum of the component earth
currents flows through the generator affected by
the fault. The cable network connected to the
generators is decisive for the fault current magni-
tude.

Example: 4.4.2 Note

In the case of 10 kV cables (lead sheath, polymer- In industry, busbar systems are designed with high
insulated) the capacitive earth-fault current lies or low resitive switchable star-point resistors. For
between 1.2 to 3.5 A/km. If with a full displace- earth-fault detection, the star-point current and
ment voltage we assume an earth current of the summation current are measured by the core-
max. 3 A and aim for a protected zone of 80 %, balance current transformer and fed into the pro-
approximately 0.6 A flows on the primary side. tection relay as a current difference (see Fig. 4).
This current (secondary approximately 10 mA) The earth-current component coming from the
can be handled reliably by the protection. star-point resistor, as well as any from the system,
contribute to the total earth-current. In order to
If the capacitive current is not sufficient where
rule out overfunction as a result of transformer
higher power levels are concerned, it is worth in-
faults, the displacement voltage serves for tripping.
vesting in an earthing transformer on the busbar
The protection then decides on generator earth-
or in disconnectable load resistors on the genera-
fault if both of the following criteria apply:
tor star point. The earth-current increases as a
result of the resistive current. n Displacement voltage is greater than setting
value U0>,
n Earth-fault current difference DIE greater than
setting value 3 I0>, magnitude.
SES = Stator earth-fault protection
Fig. 5 Earth-fault protection by differentiation with core-balance current transformers

The pickup value should be at least twice the oper- n Steam turbines: PReverse/SN ≈ 1 % to 3%
ational asymmetries. A value of 10% of the full n Gas turbines: PReverse/SN ≈ 3 % to 3%
displacement voltage is normal. n Diesel drives: PReverse/SN > 5 %
4.5 Sensitive earth-fault detection (ANSI 50/51 GN)/ However, it is advisible to measure the reverse
rotor earth-fault protection (ANSI 64R) power with the protection itself in the primary
Sensitive earth-current protection is used for de- test. About 0.5 times of the measured motoring
tecting earth faults in isolated or high-resistance energy is chosen as a setting value. The motoring
earthed systems. This protection function can also energy value can be found at the “percentage op -
serve to detect rotor winding earth-faults if the erational measured-values”.
rotor circuit is artificially displaced with a system-
frequency voltage to earth (UV ≈ 42 V by means of 4.7 Frequency protection (ANSI 81)
7XR61 coupling device). In this case the maxi- Frequency protection detects overfrequencies and
mum flowing earth current is limited by the mag- underfrequencies of the generator. If the fre-
nitude of the selected URE voltage and by the quency lies outside the permitted range, the ap-
capacitive coupling to the rotor circuit. Monitor- propriate switching operations are initiated, such
ing of the measuring circuit is provided (for this as separating the generator from the system. De-
application) as rotor earth-fault protection via the crease of frequency is caused by an increase active
sensitive earth-current measuring input. It is re- power demand the system or by malfunctions in
garded as closed if the earth current (which also the frequency or speed control. Frequency de-
flows with healthy insulation) resulting from the crease protection is also used on generators that
earth capacitance of the rotor circuit exceeds a (temporarily) feed a separate island system, since
parametrizable minimum value IEE<. Should the in such a case the reverse-power protection can-
earth-current fall below this value, a failure signal not work if the drive power fails. The generator
is issued after a short delay time (2 s). can be disconnected from the system by the fre-
quency decrease protection. Frequency increase is
A typical pickup value is approximately 2 mA. If caused for example by load shedding (separate
this value is set at 0, the monitoring stage is ineffec- island system) or malfunctions in the frequency
tive. This can become necessary if the earth capaci- control. In such cases there is a danger of self-exci-
tances are too low. The setting of the earth-fault tation of generators which feed long, no-load
pickup IEE> is selected in such a way that the insula- lines. The frequency values are generally set in ac-
tion (earth) resistances RE can be detected in the cordance with the specifications of the system or
range from about 3 kΩ to 5 kΩ: The value set power station operator. Frequency decrease pro-
should in this case be at least twice as high as the tection has the task of securing power for the sta-
interference current owing to the earth capacitances tion-service equipment by disconnecting it from
of the rotor circuit. The tripping delays the system in good time. The turbo regulator then
T IEE> and T IEE>> do not include operating times. adjusts the machine set to rated speed so that the
station-service power can continue to be supplied
4.6 Reverse-power protection (ANSI 32R) at rated frequency. A frequency increase can occur
Reverse-power protection serves to protect a tur- for example in the event of load shedding or speed
bine generator unit if, in the event of drive power control malfunction (e.g. in a separate island sys-
failure, the synchronous generator runs as a mo- tem). The frequency increase protection is thus
tor and drives the turbine and is thereby drawing used for example as overspeed protection.
the required motoring energy out of the system.
This state will endanger the turbine blades and
must be interrupted without delay by opening the Stage Cause Setting values
network circuit-breaker. For the generator there at at Delay
exists the additional danger that in the event of re- fN = 50 Hz fN = 60 Hz
sidual steam leakage (defective seal valves) after f1 Disconnection 48.00 Hz 58.00 Hz 1s
opening of the circuit-breaker, the turbine genera- from system
tor unit can be run up to overspeed. For this rea- f2 Shutdown 47.00 Hz 57.00 Hz 6s
son disconnection from the power system should f3 Alarm 49.50 Hz 59.50 Hz 20 s
only take place after detection of active power in-
f4 Alarm or 52.00 Hz 62.00 Hz 10 s
put into the machine. The value of the consumed tripping
active power is determined by the friction losses to
Setting example
be overcome and, depending on the system, is
approximately:

4.8 Overvoltage protection (ANSI 59) The tripping characteristics of the underexcitation
Overvoltage protection serves to protect the elec- protection are composed of straight lines in the
tric machine and the connected system compo- diagram, each defined by its conductance section
nents from impermissible voltage increases, 1/xd (= coordinate admittance distance) and its
thereby protecting the insulation from damage. angle of inclination α (Fig. 7).
Voltage increases result for example from incor-
rect operation in manual control of the excitation
system, from malfunction of the automatic volt-
age regulator or from (full) shedding of a genera-
tor load, separation of a generator from the sys-
tem or in separate island operation. Setting of the
limit values and delay times of the overvoltage
protection depends on the speed with which the
voltage regulator can control voltage changes. The
protection may not intervene in the control pro-
cess when it is operating trouble-free. The two-
stage characteristic must therefore always be
above the voltage time characteristic of the control
process. The long-time stage should intervene in
the event of steady-state overvoltages. It is set to
approximately 110 % to 115 % of UN and, de-
pending on the regulator speed, at 1.5 s to 5 s. Fig. 7 Underexcitation protection characteristics in the
admittance plane
4.9 Underexcitation protection (ANSI 40)

The straight lines (1/xd Char. 1) / α 1 (characteristic
Underexcitation protection (loss-of-field) protects 1) and (1/xd Char. 2) / α 2 (characteristic 2) form
a synchronous machine from loss of synchronism the static underexcitation limit. (1/xd Char.1) cor-
in the event of malfunction of excitation or con- responds to the reciprocal value of the reference
trol and from local rotor overheating. synchronous direct reactance
In order to detect underexcitation, the relay pro-
cesses all three phase currents and all three volt- 1 1 UN
= ⋅
ages as stator circuit criteria as well as the signal of x d x d 3 ⋅ IN
an external excitation voltage monitor as rotor
circuit criterion (Fig. 6). If the synchronous machine voltage regulator in-
cludes underexcitation limitation, the static char-
acteristics are set such that intervention by the
underexcitation limitation is enabled before the
characteristic 1 is reached. The generator perfor-
mance diagram can be used as a basis for setting.
If the axis sizes are divided by the rated apparent
power, the generator diagram is obtained in “per
unit” form (corresponding to a “per unit” repre -
sentation of the admittance diagram). Multiplying
1/xd by a safety factor of approximately 1.05 pro-
duces the setting value.
Fig. 6 Admittance diagram of turbo generators

For α 1 the angle of the voltage regulator under-
excitation limitation is selected, or the inclination
angle from the restraint characteristic of the ma-
chine can be read. α 1 is normally between 60° to
80°. For low active power levels, the machine
manufacturer normally specifies a minimum exci-
tation. Here the characteristic 1 is cut off from
characteristic 2 when the active power is low. α 2
is set to 90 °. With characteristic 3, the protection
can be matched to the dynamic stability limits of
the machine. If no more precise details are avail-
able, a value roughly between the synchronous di-
rect-axis reactance xd and the transient reactance
xd is selected; it should however be greater than 1.

For the angle α 3, 80 ° to 110 ° is normally se - n PROFIBUS DP, RS485 or optical 820 nm
lected, in order to ensure that only dynamic insta- double-ring ST connector,
bility can lead to tripping with characteristic 3. If n IEC 60870–5–103,
the static limit curve (consisting of characteristics n DNP3.0; RS485 or optical 820 nm double-ring
1 and 2) is exceeded, initially the voltage regulator ST connector and
must be given the opportunity to increase the ex- n MODBUS; RS485 or optical 820 nm
citation; for this reason an alarm signal is delayed double-ring ST connector
“long time” (at least 10 s). If the relay is neverthe -
less “informed” of excitation voltage failure (by an 7UM6 supports the widely used, internationally
external excitation voltage monitor via binary standardized open communication standards.
input), disconnection can take place with a short n 6. Summary
delay time.
Based on the recommendations for protection
Characteristic 1 and 2 Instantaneous Excitation signal functions [1] it has been described how, despite
steady-state stability Exc < Exc the cost aspects that have to be taken into account
Characteristic 1 and 2 Long time-delay Trippings in small-scale power generating plants, modern
steady-state stability T Char. 1 = T Char. 2 ≈ 10 s Exc < Char. 1 TRIP / relays can be used to create technically effective
Err < Char. 2 TRIP yet uncomplicated concepts.
Characteristic 1 and 2 Short time-delay Tripping
Excitation voltage failure T SHORT Uex < ≈ 1.5 s Exc < UPU < TRIP In contrast to traditional individual relays,
state-of-the-art multifunctional numerical protec-
Characteristic 3 Short time-delay Tripping
Dynamic stability T Char. 3 ≈ 0,5 s Exc < Char. 3 TRIP tion equipment now provides a wider scope of
functions. Self-monitoring contributes to avoid-
Setting of underexcitation protection
ance of underfunctions (failure to detect relay fail-
ure). A generator can be adequately protected
Note: with a single relay. For more detailed information
Selecting very short delay times can lead to on selecting functions and settings, the 7UM61
dynamic transients (possibly overfunctions). manual is recommended, chapter 2.1 of which has
It is therefore advisable not to set the times below been provided as an application handbook.
0.05 s.
4.10 Negative-sequence protection (ANSI 46) n 7. References
Negative-sequence (or unbalanced load) protec- Herrmann, H.-J.:
tion is used to detect asymmetrical loading of "Digitale Schutztechnik" (Digital Protection
three-phase induction machines. Asymmetrical Technology).
loads create a reverse field, which affects the rotors Basic principles, software, examples of
with double the frequency. Eddy currents are in- implementation.
duced on the surface of the rotor, leading to local VDE-Verlag GmbH, Berlin 1997, ISBN
overheating in the rotor end zones and slot wedges. 3-8007-1850-2.
Furthermore, interruptions, faults or incorrectly Herrmann, H.-J.: "Elektrischer Schutz von
inter-changed connections to the current trans- Kleinkraftwerken"
formers can also be detected with this protection (Electrical Protection of Small Power Stations).
function. Additionally, single and two-phase faults "Elektrizitätswirtschaft Jg."
with fault currents lower than the maximum load (Electricity Industry Year) 97 (1998) Issue 24
currents can be identified.
Siemens AG; PTD:
Setting example: SIPROTEC 7UM61 V4.1 Multifunctional
Setting value I2 permissible = 11 % ⋅ (483 A/500 A) = Machine Protection Manual
10.6 %
Factor k = 18.7 s
T cooling = 1650 s
n 5. Communication
The SIPROTEC 7UM6 relays fully satisfy the re-
quirements of modern communication technol-
ogy. They have interfaces that enable integration
into
– superordinate control centers,
– convenient parameter assignment and operation
via PC (locally or via modem connection).

System Solutions for

Protecting Medium and
Large Power Station Units
n 1. Introduction
Electrical protection is essential for reliable opera-
tion and high availability in power stations. Elec-
trical protection equipment cannot prevent faults
from occurring in the power station unit itself, but
can limit the damage and thereby shorten the
station's downtime.
Protection equipment will not be discussed in de-
tail here. Detailed reports on protection functions
and measurement procedures can be found in
“Protection of a Medium-Sized Generator up to
5 MW” and “Protection of Medium-Sized and
Large Generators with SIPROTEC 7UM6.
The subject of this publication is the design of
protection systems with regard to reliability, avail-
ability and operational safety. Various equipment
technologies (and how these affect the design and
operation of a protection system) are compared.
n 2. Safety and availability of the protection

system
Protection equipment serves to detect faults or Fig. 1 SIPROTEC Generator protection for power stations
impermissible operating conditions in power sup-
ply systems and to shut them down when such Possible failures in electromechanical protection
conditions occur. When the station is operating relay can go unnoticed and the relay then almost
trouble-free, conventional protection equipment always fails to function properly. Availability of
does not indicate whether it is functioning cor- the protection system is no longer ensured. For
rectly or not. The operator must therefore check this reason the protection equipment for large-
to make sure that the protection relays are work- scale plants – including of course large power
ing, using recurrent function tests. This confirma- station units – is duplicated. The statistical proba -
tion of correct operation only applies for the du- bility that both protection relays will fail at the
ration of the function test. For the time between same time is so slight that the availability of a re-
such periodic function tests, no dependable state- dundant protection system can be considered as
ment about the condition of the protection relays sufficient (see Fig. 2).
can be made. Maintenance and testing is discussed
in more detail in Chapter 4 of this application
example.

Safety, availability System design Underfunction Overfunction

1-out-of-1 5 years 5 years
Electromechanical protection
1-out-of-2 600 years 2,5 years
Cause of fault Relay failure Transient effects
2-out-of-2 2,5 years 21,750 years
2-out-of-3 200 years 7,250 years
Result Underfunction Overfunction
Table 1 Overview system design
Countermeasure Redundancy None The MTBF (Mean Time Between Failures) was
Fig. 2 Safety design for electromechanical protection calculated for different system configurations. The
2-out-of-3 system in analog design offers the
In the use of analog electronic protection relays an highest reliability against overfunction and under-
additional aspect comes into play. A fault on an function. However, a 2-out-of-3 system is techni-
electromechanical protection relay almost always cally very complex and cost-intensive and for this
causes relay failure. A relay fault in an analog sta- reason only used in a few nuclear power plants.
tic protection relay can cause either underfunction A notable characteristic of numerical protection
or overfunction, with approximately the same relays are continuous self-monitoring of hardware
probability. Underfunction can, as with electro- and software. Reliability against overfunction and
mechanical protection systems, be mastered by re- underfunction is inherently provided (see Fig. 4).
dundant design. The danger of overfunction can Any relay failure causes blocking of individual
be prevented in a limited sense by dual-channel protection functions or of the entire relay. This
design of the measuring circuits (see Fig. 3). provides an effective measure against
overfunction of the protection relays and thereby
Safety, availability of the protection system as a whole. Relay failure
is signaled at the same time. This feature means
Analog-static protection that defective protection relays can be immedi-
Cause of fault Relay failure Transient effects ately replaced and the statistical availability of the
protection system is enhanced.
In power stations that either have a relatively
Result Underfunction Overfunction Overfunction low-output or are of comparatively minor signifi-
cance for secure power supplies, the requirement
Countermeasure Redundancy Dual-channel Measuring for a redundant protection system can conse-
electronic repetition quently be re-assessed. If brief shutdown of the
circuit design (limited) plant is acceptable, the investment costs can be re-
duced by dispensing with redundancy. “Brief ”
here means a period of one or two days until the
2-out-of-3 replacement equipment is installed and put into
Fig. 3 Safety design for analog-static protection operation.
In most power stations, however, shutdown due
If a high level of safety and availability is to be at- to defective protection relays is not acceptable nor
tained with analog technology, the 2-out-of-3 is continued operation without complete protec-
principle can be applied. Three identical or equiv- tion. Consequently, a redundant protection sys-
alent protection systems must be linked with each tem must be considered in medium and large
other externally so that two tripping signals from power station units. From a technical point of
different systems are always connected in series. view, complete redundancy of the protection
This way a very high level of safety from functions enables continuous operation of the
overfunction and underfunction can be achieved. unit for a short time until defective protection re-
In a study in the 1970s the safety and reliability of lays are replaced.
different analog protection system configurations
were statistically researched (see Table 1).

For the sake of completeness, the influence of

transient phenomena on the behavior of protec- Safety, availability
tion relays (Figs. 2, 3 and 4) must be mentioned. Numerical protection
Electromechanical protection relays offered, apart
from their time-lag characteristics, practically no Cause of fault Relay failure Transient effects
effective compensation for transient influences.
With analog technology, disturbing transient
measured variables can be eliminated to a certain Result Underfunction Overfunction Overfunction
extent by multiple measurements. Only numerical
technology enables reliable control of transient Countermeasure Self-monitoring Self-monitoring Measuring
disturbances through consistent digital filtering with with repetition
and repeat measurements. failure alarm + relay blocking (consequent)
redundancy
Producing complete redundancy does not require
Fig. 4 Safety design for numerical protection
100% duplication of all protection relays. Redun-
dancy can also be achieved by two different mea-
surement procedures for one and the same fault.
For example a redundant protection concept can
be realized against short-circuits by combining
current differential protection and impedance
protection in mutually independent relays. For
some protection functions, differing measurement
principles are even desirable. Duplicate differen-
tial protection provides two fast and selective
protection against short-circuits in the machine.
Impedance protection as the second stage of
short-circuit protection includes at the same time
backup protection against power system faults
(see Figs. 5 and 6).
A few special features must be noted in the redun-
dant design of protection functions whose princi-
ple is based on supplying an external voltage Fig. 5 Diverse redundancy: Backup protection against system faults
(100% stator earth-fault protection with 20 Hz in-
jection and rotor earth-fault protection). The
auxiliary devices cannot be operated in duplicate
on the generator. It is, however, possible and ap-
propriate to operate the protection functions
themselves in redundancy. Here, the measuring
inputs of both protection relays are fed in parallel
from the same 20 Hz or 1 Hz frequency generator.
If a very high level of statistical availability is
aimed for, the auxiliary devices can be installed in
duplicate with one changeover switch in the pro-
tection cabinet. If any auxiliary devices fails, the
parallel relay is activated by the changeover
switch.
Fig. 6 Mirrored redundancy: No backup protection against system faults

n 3. Indication processing and communication also be requested via the bus connection as neces-
Numerical protection relays offer the operator sary. An information network with connected PC
considerably more scope for status and fault is available for detailed fault analysis after tripping
indications than do conventional relays. In the in- on faults has occurred. Via this means of commu-
terests of more reliable operation and control of nication, the protection expert can read out all the
the power station, it is the planning engineer's re- available information from the protection relays.
sponsibility, from the abundance of available in- Using the message lists and the transient fault re-
formation, to provide (to operating personnel) cords the expert can draw up an exact profile of
precisely those indications and measured values the fault that led to the power station unit being
that are needed for each task. Rather than suc- shut down. Alternatively, this detailed informa-
cumbing to the temptation to make all available tion can be read out on the front of the protection
information accessible at each workstation, the relays.
planning engineer is responsible for working out
an intelligent indication signalling concept. n 4. Maintenance and testing
Continuous self-monitoring of numerical protec-
This means that each member of staff receives the
tion relays creates new opportunities for opera-
information needed to make quick and reliable
tion and testing. While it was essential to monitor
decisions about technical operation and control of
the state of conventional protection relays by
the power station. Fig. 7 provides a concept for
means of periodic function tests, numerical pro-
such an information network. Group indications
tection relays do most of this work themselves.
from the protection system are provided in the
Fig. 7 Integration of protection relays into power station control and protection system
power station control room. These indications en- The self-monitoring installed in each numerical
able a quick overview of the operating state of the protection relay continuously checks that the
power station unit in terms of electrical faults or hardware and software are working correctly. This
impermissible operating states. The recommended produces a series of consequences for mainten-
spontaneous indications for the control room are: ance and testing of a numerical protection system.
n Tripping on faults
n Protection faulty
n Negative-sequence (unbalanced load) alarm
n Stator earth-fault alarm
n Rotor earth-fault alarm
n Underexcitation alarm
In addition to these spontaneous indications,

measured values from the protection relays can

Protection relays can be regularly checked by feed-

Tasks: – Detection of a wire breakage in the voltage circuit
ing in fault currents and voltage over a longer pe- - Detection of device-internal faults in the voltage circuit
riod of time. Such overall testing should be carried
U min
out in the context of regular power station main- Operating condition: < SYM.FAK.U und U max ≥ k ⋅ SYM.UGREN
U max
tenance. Since this check does not take place dur-
ing operation of the power station, special test
switches or plugs are not necessary. Current and
voltage is fed in via the cabinet terminals. In view
of the multi-functional protection relay concept
for generators it is practically impossible to test
individual protection functions without interven-
ing in the relay parameterization. This is also not
necessary because all protection functions of a
relay are handled on the same hardware. The con-
ventional approach involving protection charac-
teristics with measurement of tolerances is not
provided in numerical protection. Thanks to digi-
tal measured-value processing, the ageing or tem-
perature drift of analog components is nowadays
practically unknown. The function test is thus Fig. 9 Voltage balance monitoring
limited to a sensitivity check of the protection
relay as a whole. As a reaction to an identified fault, a defective
protection relay produces either just an indication
During power station operation the self-monitor-
or blocks itself partly or wholly to prevent over-
ing performs continuous testing of the numerical
function. Taking into account the recognized
protection relays. As well as monitoring the pro-
fault, the protection relay reacts in a graded way
gram flow using Watchdog, the correct condition
according to the seriousness of the fault (Fig. 10).
of the hardware components is continually
checked. This takes place for example using write/
read cycles for the memory and with the process-
ing of reference parameters for the analog/digital
converter. The protection relay also monitors ex-
ternal connections, where possible. One example
of many is monitoring of the measuring voltage
balance (Figs. 8 and 9).
Fig. 8 Voltage balance monitoring

Possible reactions of a monitoring function
Minor faults n Transmitting an indication

n Blocking individual protection functions. Only functions relating to the
fault are blocked
n Withdrawal from service of all protection functions.
The relay remains operable and indications can still be read out.
n Restarting the equipment. A maximum of 2 restarts are possible.
After the 3rd restart the relay switches to monitor mode.
n Switch to monitor mode. This usually means a hardware fault.
The cause can be determined by evaluating a fault buffer
Serious faults
Fig. 10 Design of monitoring functions
During normal power station operation, numeri- n 5. Summary

cal protection relays make it possible to check the The development of numerical technology for
complete measured-value acquisition circuit by protection relays brought about considerable ben-
reading out status measured-values. Measured efits in terms of enhanced performance, reliability
values can be indicated on the equipment display and availability of the protection system. These
without adversely affecting the protection func- benefits upgrade the availability of the power sup-
tions. A comparison of these measured values ply system. Quantity and quality of information
with other measuring devices or protection relays retrieved from the numerical protection relays
represents an inspection and test sequence which have reached such a high level that operation and
would be unthinkable in conventional technology maintenance of the power station can be per-
without additional equipment. This test step formed much safer and with less efforts. The in-
includes the following power station and device crease in functionality in the numerical protection
components: relays requires a well-designed protection and in-
n Current and voltage transformer formation system, making the best of the many
different opportunities.
n Transformer supply cables
n Cubicle wiring for the measuring circuits
n Input measuring transmitter for the protection
relays
n Analog/digital converter
n Measured-value memory
This inspection and test sequence is of minimal
duration and can be carried out at any time, with-
out intervening in the protection processing of the
relay.

Protection of Medium-
Sized and Large Generators
with SIPROTEC 7UM6
n 1. Introduction
Medium-sized and large generators make a major
contribution to power generation. They carry the
basic load and ensure the stability of an energy
system.
The task of electrical protection in these systems is
to detect deviations from the normal condition
and to react according to the protection concept
and the setting. Based on experience with larger
power station units, cost-effective protection con-
cepts can also be implemented with SIPROTEC
relays for medium-sized generators.
The scope of protection must be in reasonable re-
lation to the total system costs and the importance
of the system.
n 2. Basic connections
In medium-sized and large power stations the
generators are operated exclusively in unit
connection.
In the unit connection the generator is linked to
the busbar of the higher voltage level via a trans-
former. In the case of several parallel units, the
generators are electrically isolated by the trans-
formers. A circuit-breaker can be connected be- Fig. 1 SIPROTEC generator protection
tween the generator and the transformer
(see Figs. 2 and 3).
Fig. 2 Block diagram of generator protection

Fig. 3 Redundant protection concept for large generators
n 3. Protection concept Full redundancy (see Fig. 5)

Components of the protection concept are: In this system design the redundancy concept is
n The redundancy concept applied consistently throughout by duplicating all
the essential components. As shown in Fig. 5, the
n The tripping concept
redundancy begins with separate transformers or
n Protection function scope transformer cores, continues through the protec-
tion relays, and the TRIP signal is passed through
3.1 Redundancy concept separate DC voltage paths to switchgear with 2
The redundancy concept is crucial in the design of circuit-breaker coils (see Fig. 5). In the protection
protection systems. Many concepts are based on relays the protection functions can be duplicated
the n-1 principle. That means that the failure of a on the one hand; on the other hand additional
component is under control and does not lead to a protection functions with different measuring
total system failure. However, this principle is not principles are desired. Typical examples are
always applied consistently. In smaller systems earth-fault and short-circuit protection.
there is a compromise between redundancy and
costs. The following strategies are common in The displacement voltage measurement covers
practice for medium-sized and large generators. about 90 % of the protected zone in the event of
an earth fault. The totally different method –
Partial redundancy (see Fig. 4) injecting an external voltage (20 Hz) into the
At least 2 protection relays are used here. The pro- stator circuit – ensures 100 % protection.
tection relays/functions are selected so that the
The same can be implemented for the short-
system can continue to operate when a relay fails.
circuit protection. The main protection is the fast
However, certain restrictions have to be accepted.
and selective current differential protection. Sup-
This system design is seldom used in high-power
plementary to this, impedance protection is used,
generators. The protection is connected to the
with which the backup protection for the power
same transformers for example.
system protection can be achieved by appropriate
grading.

Fig. 4 Example: Partial redundancy Fig. 5 Example: Full redundancy
3.2 Tripping concept

The special feature of generator protection is that l Generator circuit-breaker
different switching devices have to be activated de- l De-excitation switch
pending on the fault. The number is basically de-
l Turbine valve closing
l System switch
termined by the system/plant concept. As a rule,
l Auxiliaries switch 1
most of the switching devices need to be actuated l Auxiliaries switch 2
in the larger units. Special trippings are used in l Auxiliaries transfer
hydro-electric power stations. l Spray water system, unit transformer
Fig. 6 shows the basic concepts. On one side there l Spray water system, station-service trans-
former
are the switching devices to be actuated and on the
l Backup
other side the connected protection functions. l Special trippings for hydropower
The tripping program or tripping concept de- (e.g. braking)
pends on the recommendations/experiences and Fig. 6 Protection tripping by the matrix
the operating conditions. There are two opposing
philosophies. The tripping program is determined
The function matrix is scalable to meet different re-
individually by a tripping matrix (a software ma-
quirements (see Table 1).
trix in digital technology) and the switching de-
vices are activated directly. The other (American-
The selection simplifies division into
influenced) variant reduces the tripping to two
object and application-related groups.
programs, e.g. exclusive shutdown of the gener-
ator and shutdown of the power station unit.
Lockout relays are used to control the switching
devices. The protection needs only a few trip con-
tacts.
3.3 Protection function scope

Numerous protection functions are necessary for
reliable protection of electrical generators. The
scope and the combination are determined by
various factors such as generator size, operating
principle, system design, availability requirements,
experiences and philosophies. This automatically
leads to a multifunctionality which can be con-
trolled excellently by numerical technology.

Protection functions Generator rated power n 4. Protection functions and setting

The basic connection (Fig. 2) is considered for the
5 - 50 MVA 50 - 200 MVA > 200 MVA
setting value calculation – with generator data
Stator earth-fault protection 90 % n n n from the Table 3. Manufacturer characteristics
Stator earth-fault protection 100 % n n (e.g. power diagram) are necessary for some pro-
tection settings. The physical backgrounds and the
Differential protection n n n
formulae for calculation are in the manual. The
Overcurrent-time protection n £ £ secondary setting values are shown.
Impedance protection n n
Rotor earth-fault protection n n n
Negative-sequence n n n
(or load unbalance) protection
Underexcitation protection n n n
Out-of-step protection £ n
Stator overload protection n n n
Rotor overload protection n
Overvoltage protection n n n
Frequency protection f > n n n
Frequency protection f < n n n
Reverse-power protection n n n
Undervoltage protection ¤ ¤ ¤
Overexcitation protection £ n n
n Available
£ Optional
¤ Pumped-storage station (motor protection and phase modifier operation)
Table 1 Recommended protection functions according to generator rated power
A function selection taking redundancy into

consideration is shown in Table 2.
Protection group A Protection group B

(System 1, 7UM622) (System 2, 7UM622)
Stator earth-fault 100 % Stator earth-fault 90 %
Differential protection Differential protection (as
unit protection)
Impedance Impedance
Rotor earth-fault Negative-sequence
Negative-sequence Underexcitation
Underexcitation Out-of-step
Overvoltage Stator overload
Frequency f >< Overvoltage
Reverse power Frequency f ><
Overexcitation Reverse power
Overexcitation
Table 2 Function selection for a redundancy concept

Generator data
Rated voltage UN 15.75 kV ± 5 %
Rated apparent power (40 °C cold gas) SN 327 MVA
Circuit-breaker cos ϕ 0.8
Rated active power PN 261.6 MW
Rated current IN 12 kA
Rated frequency fN 50 Hz
Maximum overexcitation (U/f)max % from the manufacturer’s overexcitation characteristic
Permissible overexcitation duration t (U/f)max from the manufacturer’s overexcitation characteristic
Synchronous longitudinal reactance xd 264.6 %
(for drum rotor generators: xd = xq)
Transient reactance xd1 29.2 %
Maximum exciter voltage Uexc-min 77 V
Maximum continuous permissible inverse current Imax prim / IN 10 %
Thermal continuous permissible primary current Imax / IN 1.2
2
Asymmetry factor (I2) K = (I2/IN) t 20 s
Current transformer Iprim Isec ü Target

Star-point side
T1, core 1 14 kA 1A 14 000 System 1
Busbar side
110 kV side
T3, core 1 2 000 A 1A 2 000 System 2
Earthing transformer Uprim Usec ü Target

T4; U0 15.75 kV/ 3 5 V/ 3 54.56 System 1
External voltage Uprim Usec ü Target

transformer
Generator side
T5, UL1, UL2, UL3 15.75 kV/ 3 100 V/ 3 157.5 System 1, 2
Unit transformer data

Vector group Ynd5
Total coupling capacity HV-LV Ck 14.4 nF (4.8 per phase)
Maximum overexcitation (U/f)max 120 %
Permissible overexcitation time t (U/f)max from the manufacturer’s overexcitation characteristic
Permissible overload Imax / IN from the manufacturer’s overexcitation characteristic
Winding Primary Secondary
Rated voltage UN 115 kV 15.75 kV
Rated apparent power SN 318 MVA 318 MVA
Rated current IN 1.596 kA 11.657 kA
Short-circuit voltage uSC 15 %
Control range of the tap changer ± 9 x 1.25 %
Table 3 Data of the power station unit with gas turbine

4.1 Current differential protection 4.2 Stator overload protection

(ANSI 87G, 87M, 87T) (ANSI 49)
The function is the instantaneous short-circuit The overload protection should protect the stator
protection in generators, motors and transformers winding of generators and motors against exces-
and is based on the current differential protection sively high continuous current overloads. All load
principle (node set). The difference and restraint cycles are evaluated by a mathematical model. The
(stabilization) current is calculated from the phase basis for the calculation is the thermal effect of the
currents. Optimized digital filters safely attenuate current r.m.s. value. The transformation corre-
disturbance variables such as aperiodic DC ele- sponds to IEC 60255-8.
ments and harmonics. The high resolution of the
measuring variables enables small difference cur-
rents (10 % of IN) to be picked up, i.e. a very high The cooling time constant is prolonged automati-
sensitivity. A settable restraint characteristic al- cally, dependent on the current. If the ambient
lows optimum adaptation to the conditions of the temperature or coolant temperature is fed in
protected object. through a transducer (MU2) or via the
PROFIBUS-DP, the model automatically adapts
Setting instructions to the ambient conditions; otherwise a constant
An important setting is the position of the star ambient temperature is assumed.
points of the current transformer sets on both The following table shows the setting options and
sides of the protected object. In addition, the rated the setting example of important parameters
data (SN GEN/MOTOR, UN GEN/MOTOR) of the genera- (without taking the ambient or coolant tempera-
tor to be protected and the primary and secondary ture into account).
rated currents of the main current transformers
are requested on both sides. The setting values Parameter Setting options Setting
refer to these. In addition, they are used for exam- k-factor 0.1 to 4.0 1.11
ple to determine the primary measured values.
Thermal warning 70 to 100 % 95 %
As an additional security measure against unwant- level
ed operation when connecting a previously non- Current warning 0.1 to 4.0 A 1.0 A
energized protected object, the increased pickup level
value can be switched on when starting up. kt-time factor at 1.0 to 10.0 1.0
standstill
The table below shows the setting options of
selected parameters. The settings are relevant for Limit current for 0.5 to 8.0 A 3.30 A
the generator and not for the unit (protection the thermal replica
group A). Dropout time after 10 to 15000 s 100 s
emergency start
Parameter Setting options Default *) Table 5 Parameter overview for the stator overload
protection
Pickup value of the trip 0.05 to 2.0 I/INObject 0.2 I/INObject
stage Idiff> The setting ranges and presettings (defaults) are
Delay of the trip stage 0 to 60.0 s; ∞ 0.00 s specified for a secondary rated current of IN = 1 A.
Idiff> At a secondary rated current of IN = 5 A, these va-
Pickup value of the trip 0.05 to 12.0 I/INObject 7 I/INObject lues must be multiplied by 5. The ratio of the cur-
stage Idiff>> rent transformer must be taken into account
Delay of the trip stage 0 to 60.0 s; ∞ 0.00 s additionally for settings of primary values.
Idiff>>
Slope 1 of the trip 0.1 to 0.5 0.15
characteristic
Foot of slope 1 of the 0 to 2.0 I/INObject 0 I/INObject
trip characteristic
Slope 2 of the trip 0.25 to 0.95 0.5
characteristic
Foot for slope 2 of the 0 to 10.0 I/INObject 2.50 I/INObject
trip characteristic
Table 4 Parameter overview for the differential
protection
*) In this example, most of the default settings can be used.

4.3 Negative-sequence protection (ANSI 46) Parameter Setting options Setting

Asymmetrical current loads of the three phases of
Continuously permissible 3.0 to 30.0 % 8.6 %
a generator lead to heating up in the rotor due to load unbalance
the reverse field. The protection detects an unbal-
Delay time of warning 0 to 60.0 s; ∞ 10.0 s
anced load of three-phase generators. It operates stage
on the basis of symmetrical components and eval-
uates the negative-sequence component of the Asymmetry factor K 2.0 to 100.0 s; ∞ 11 s
phase currents. The thermal processes are taken Cooling time of thermal 0 to 50000 s 1500 s
into account in the algorithm and lead to an model
inverse-time characteristic. In addition, the nega- Excitation current I2>> 10 to 100 % 51.4 %
tive-sequence is evaluated by a definite-time
Delay time T I2>> 0 to 60.0 s; ∞ 3.0 s
warning and tripping stage which is supplemented
by delay elements. Table 6 Parameter overview for negative-sequence
protection
Thermal characteristic
The protection prevents damage due to out-of-
The generator manufacturers specify the permissi- steps resulting from underexcitation. The complex
ble negative-sequence with the following formula: master value is calculated from the generator ter-
minal voltage and current. The protection func-
K
t perm = 2 tion offers three characteristics for monitoring the
 I2  static and dynamic stability. The exciter voltage
 
 IN  can be fed in through a transducer and a fast re-
sponse of the protection can be achieved by timer
tperm = Maximum permissible application time of switching in the event of a failure. The straight
the negative-sequence current I2 line characteristics allow optimum adaptation of
K = Asymmetry factor (generator constant) the protection to the generator diagram. The set-
ting values can be read out directly from the per-
I2/IN = Negative-sequence (ratio of negative- unit representation of the diagram. The positive-
sequence current I2 to rated current IN) sequence components of the currents and voltages
The asymmetry factor is generator-dependent and are used for calculating the variables, whereby
represents the time in seconds for which the gen- correct operation is ensured even under asymme-
erator may be loaded at the maximum with 100 % trical conditions.
unbalanced load. The factor is mainly in the range Setting instructions
between 5 s and 30 s. On exceeding the permissi- The tripping characteristics of the underexcitation
ble load unbalance (value of the continuously per- protection are made up of straight lines in the
missible negative-sequence current), simulation of master value diagram, defined by their reactive
the heating of the object to be protected in the re- part of the admittance 1/xd and their angle of
lay begins. The current-time-area is calculated inclination α.
continuously taking different load cases correctly Table 7 shows the settings for this application
into consideration. If the current-time-area example.
((I2/IN)2· t) reaches the asymmetry factor K, trip-
ping takes place with the thermal characteristic. Parameter Setting options Default *)
Table 6 shows the setting options and the setting Pickup threshold 1/xd characteristic 1 0.25 to 3.0 0.37
example. Characteristic slope characteristic 1 50 to 120 ° 80 °
Delay time characteristic 1 0 to 60.0 s; ∞ 10.0 s
Pickup threshold 1/xd characteristic 2 0.25 to 3.0 0.33
Characteristic slope characteristic 2 50 to 120 ° 90 °
Pickup threshold 1/xd characteristic 3 0.25 to 3.0 1.0
Characteristic slope characteristic 3 50 to 120 ° 100 °
Table 7 Parameter overview for underexcitation protection

4.5 Reverse-power protection (ANSI 32R) Parameter Setting options Setting

The reverse-power protection monitors the active
power direction and picks up in the event of a me- Delay time with emer- 0 to 60.0 s; ∞ ∞s
gency tripping
chanical energy failure, because the drive energy is
then taken out of the system. This function can be Delay time without 0 to 60.0 s; ∞ 6.00 s
used for operational shutdown of the generator emergency tripping
but also prevents damage to steam turbines. The Pickup threshold re- 30.0 to 0.50 % -3.42 %
position of the emergency tripping valve is entered verse power
as binary information. This switches between two Pickup seal-in time 0 to 60.0 s; ∞ 1.00 s
delays of the open command. The reverse power is Table 8 Parameter overview of reverse-power
calculated from the positive phase sequence sys- protection
tems of current and voltage. Asymmetrical system
conditions therefore do not impair the measuring 4.6 Impedance protection (ANSI 21)
accuracy. This fast acting short-circuit protection protects
the generator or unit transformer on the one hand
If reverse power occurs, the turbo set must be dis- and is the backup protection for the system. It has
connected from the system, because operation of two adjustable impedance stages whereby the first
the turbines is not permissible without a certain stage is additionally switchable by a binary input.
minimum steam throughput (cooling effect), or The impedance measuring range can be extended
the motorized load is too great for the system in a with open system switch. The overcurrent pickup
gas turbo set. with undervoltage seal-in provides reliable pickup
The trip command is delayed by an adjustable and loop selection logic for determining the faulty
time to bridge any brief power consumption loop. It also allows correct measurement through
during synchronization or in the event of power the transformer.
swings due to system faults. In the event of a Setting instructions
closed emergency tripping valve on the other The maximum load current occurring during op-
hand, the unit must be shut down with a short de- eration is decisive for setting the overcurrent
lay. By entering the position of the emergency pickup. Pickup by overload must be ruled out.
tripping valve through a binary input, the short The pickup value must therefore be set above the
delay becomes effective in the event of emergency maximum expected (over) load current. Recom-
tripping. It is also possible to block the tripping by mended setting: 1.2 to 1.5 times rated generator
means of an external signal. current.
The value of the consumed active power is deter- The pickup logic corresponds to that of the defi-
mined by the friction losses to be overcome and, nite-time overcurrent protection I>. If the excita-
depending on the system, is approximately: tion is derived from the generator terminals and
n Steam turbines: Prev/SN 1 % to 3 % the short-circuit current is able to drop below the
n Gas turbines: Prev/SN 3 % to 5 % pickup value due to collapsing voltage, the under-
n Diesel drives: Prev/SN > 5 %
voltage seal-in is activated.
The undervoltage seal-in U< is set to a value just
However, it is advisable to measure the reverse
below the lowest phase-to-phase voltage occurring
power with the protection itself in the primary
during operation, e.g. to U< = 75 % to 80 % of the
test. About 0.5 times of the measured motoring
rated voltage. The seal-in time must be greater
energy, which can be read out under the “percent-
than the maximum fault clearance time in the
age operational measured values”, is chosen as a
backup case. (Recommended: This time + 1 s).
setting value.
Table 8 shows the setting of selected parameters.

As described in the manual, the protection has Parameter Setting options Setting
three characteristics which can be set independ-
Pickup value of the 0.10 to 20.0 A 1.20 A
ently: overcurrent pickup
n Zone (instantaneous zone Z1) with the setting Pickup voltage of the 10.0 to 125.0 V 75.0 V
parameters undervoltage seal-in
ZONE Z1 reactance = reach, Seal-in time of the 0.1 to 60.0 s 10.0 s
ZONE1 T1 = 0 or short delay if necessary. undervoltage seal-in
n Overreach zone Z1B, controlled externally by a Trip time of the end time stage 0.1 to 60.0 s 3.0 s
binary input with the setting parameters
Impedance zone Z1 0.05 to 130.0 Ω 7.28 Ω
OVERR. Z1B reactance = reach,
OVERR. T1B T1B = 0 or short delay if necessary. Trip time zone Z1 0 to 60.0 s; ∞ 0.30 s
n 2nd zone (Zone Z2) with the setting parameters Impedance overreach stage Z1B 0.05 to 65.0 Ω 11.44 Ω
ZONE Z2 reactance = reach, Trip time overreach stage Z1B 0 to 60.0 s; ∞ 8.00 s
ZONE2 T2 T2 should be chosen so high that it
is above the grading time of the system protec- Impedance zone Z2 0.05 to 65.0 Ω 11.44 Ω
tion. Trip time Z2 0 to 60.0 s; ∞ 8.00 s
n Undirected final stage with the setting parame- Table 9 Parameter overview for the impedance protection
ters T END T END must be chosen such that
the second or third stage of the series-connected 4.7 Overvoltage protection (ANSI 59)
power system distance protection is overreached. This protection prevents insulation faults as a
Since it can be assumed that the impedance pro- result of high voltage. Optionally the maximum
tection measures into the generator transformer, phase-to-phase voltages or phase-to-earth volt-
it must be ensured that the parameterization ages (in low-voltage generators) can be evaluated.
selection sufficiently considers the control range In the phase-to-phase voltages, the measuring re-
of the transformer. For ZONE Z1, a reach of sult is independent of the zero point displace-
about 70 % of the zone to be protected is therefore ments resulting from earth-faults. The protection
normally chosen (i.e. about 70 % of the trans- function is designed in two stages.
former reactance) without or with only slight de- The setting of the limit values and delay times of
lay (i.e. = 0 s to 0.50 s). the overvoltage protection depends on the speed
For ZONE Z2, the reach could be set to about at which the voltage regulator can regulate volt-
100 % of the transformer reactance or a system age fluctuations. The protection may not interve-
impedance additionally. The corresponding time ne in the regulating process of the voltage regula-
stage ZONE2 T2 must be chosen so that it tor when it is operating trouble-free. The two-
overgrades the system protection relays of the fol- stage characteristic must therefore always be
lowing lines. above the voltage time characteristic of the regu-
lating process.
The following settings apply for the configuration
example (without activation of the out-of-step Setting instructions
block): The long-time stage U> and T U> should inter-
vene in the case of steady-state overvoltages. It is
set to about 110 to 115 % UN and to 1.5 to 5 s,
depending on the regulator speed. In the event of
a full load disconnection of the generator, the
voltage first rises according to the transient volt-
age and is reduced to its rated value by the voltage
regulator afterwards. The U>> stage is generally
set as a short-time stage so that the transient pro-
cess in full load shutdown does not lead to trip-
ping. About 130 % UN – with a delay T U>> ran-
ging from zero to 0.5 s – are usual (for example)
for U>>.

Parameter Setting options Default *) 4.9 Overexcitation protection (ANSI 24)

Overexcitation protection serves to detect an im-
Pickup voltage U> 30 to 170 V 115 V
permissibly high induction (proportional to U/f)
Delay time T U> 0 to 60 s; ∞ 3s in generators or transformers which leads to
Pickup voltage U>> 30 to 170 V 130 V thermal overloading. This danger can occur in
Delay time T U>> 0 to 60 s; ∞ 0.50 s
start-up processes, in full load disconnections, in
“weak” systems and in separate island operation.
Dropout ratio RV U> 0.90 to 0.99 0.95 The inverse-time characteristic is set with the ma-
Table 10 Parameter overview for the overvoltage nufacturer data by way of 8 points. A definite-
protection time alarm stage and a short-time stage can be
used additionally. Apart from the frequency, the
4.8 Frequency protection (ANSI 81) maximum of the three phase-to-phase voltages is
Frequency protection prevents impermissible used for calculating the quotient U/f. The monito-
loading of the equipment (e.g. turbine) at under rable frequency range is between 11 and 69 Hz.
and overfrequency, and also serves often as a mo-
nitoring and control element. The function is de-
signed in four stages, whereby the stages can The overexcitation protection contains two-staged
operate either as under or overfrequency protecti- characteristics and one thermal characteristic for
on. Each stage can be delayed individually. The approximate simulation of the heating of the pro-
complex frequency measuring algorithm also fil- tected object due to overexcitation. On exceeding
ters out the fundamental harmonic reliably in the of an initial pickup threshold (alarm stage U/f), a
event of distorted voltages and determines fre- time stage T U/f > is started, at the end of which
quency very accurately. The frequency measure- an alarm message is output.
ment can be blocked by an undervoltage stage. The limit value of induction in relation to the
Setting instructions rated induction (B/BN) specified by the protected
object manufacturer forms the basis for setting
If the frequency protection is used for the task of
the limit value U/f >.
system decoupling and load shedding, the setting
values depend on the concrete system conditions. The characteristic for a Siemens standard trans-
Usually a grading according to the importance of former has been chosen as default. If the protected
the consumers or consumer groups is aimed at for object manufacturer supplies no data, the default
load shedding. Other applications are to be found standard characteristic is retained. Otherwise, any
in the power station sector. Basically the frequen- tripping characteristic can be specified by point-
cy values to be set depend on the presettings of the by-point input of parameters by a maximum of
system or power station operator. 7 straight sections.
The following table shows the settings which meet
practical requirements.
Parameter Setting Setting

options
Pickup frequency f1 40 to 65 Hz 47.5 Hz
Delay time T f1 0 to 600 s 40 s
Pickup frequency f2 40 to 65 Hz 47 Hz
Pickup frequency f3 40 to 65 Hz 51.50 Hz
Pickup frequency f4 40 to 65 Hz 52 Hz
Minimum voltage 10 to 125 V; 0 65 V
Table 11 Parameter overview for frequency protection

4.10 90 % stator earth-fault protection directio- Parameter Setting options Setting

nal, non-directional (ANSI 59N, 64G, 67G)
Pickup voltage U0> 2 to 125 V 10 V
In generators operated with an isolated star point,
an earth fault is indicated by occurrence of a dis- Pickup current 3I0> 2 to 1000 mA 5 mA
placement voltage. In unit connection, the displa- Slope angle of the 0 to 360 ° 15 °
cement voltage is a sufficient, selective protection directional lines
criterion. If a generator is connected electrically to Delay time T SES 0 to 60 s; ∞ 0.30 s
a busbar, the direction of the flowing earth-cur-
Table 12 Parameter overview for the stator earth-fault
rent must be evaluated additionally for selective protection
earth-fault detection. The protection measures the
displacement voltage on a voltage transformer in 4.11 Rotor earth-fault protection (ANSI 64R)
the generator star point, or at the open delta win- This protection function can be effected with the
ding of a voltage transformer. Optionally the 7UM62 in three ways. The simplest form is the
zero-sequence voltage can also be calculated from method of rotor earth-current measurement (see
the phase-to-earth voltages. Depending on the the manual: Chapter on sensitive earth-fault
system design, 90 to 95 % of the stator winding of protection, resistance measurement at system
a generator can be protected. frequency voltage).
When starting, it is possible to switch over to The second form is rotor earth-resistance measu-
zero-sequence voltage measurement via an exter- rement with system frequency voltage coupling
nally coupled signal. Various earth-fault protecti- into the rotor circuit (see the manual: Chapter on
on concepts can be implemented with the func- rotor earth-fault protection). The coupled voltage
tion, according to the protection setting. and the flowing rotor earth-current are detected
by the protection. Taking account of the complex
Setting instructions resistance of the coupling device (7XR61), the ro-
For generators in unit connection, the pickup tor earth resistance is calculated by a mathemati-
value must be chosen so high that displacement cal model. This form is often used for medium-
voltages, which affect the stator circuit through sized generators.
the coupling capacitances of the unit transformer,
do not lead to pickup. The damping by the load The third form is resistance measurement with
resistance must also be taken into account here. square-wave voltage injection of 1 to 3 Hz. In larg-
The setting value is twice the displacement voltage er generators a higher sensitivity is required. On
value coupled in at full system displacement. Final the one hand, disturbances caused by the rotor
specification of the setting comes during commis- earth capacitance must be eliminated better, and
sioning with primary variables according to the on the other hand the interference margin to the
manual. Tripping in the event of stator earth-fault harmonic (e.g. 6th harmonic) of the exciter device
is set time-delayed (T SES). The overload capacity must be increased. The injection of a low-frequen-
of the load equipment must also be taken into ac- cy square-wave voltage into the rotor circuit has
count when setting the delay. All set times are ad- proven effective here (recommended for this ap-
ditional delays which do not include the operating plication). The square-wave voltage injected by
times (measuring time, dropout time) of the pro- the controller unit 7XT1 leads to polarity reversal
tection function. of the rotor earth capacitance. Through a shunt in
the 7XT1, the flowing earth-current is detected
Table 12 shows the setting options for selected and injected into the protection (measuring in-
parameters. The settings are selected for this pro- put). In a fault-free case (RE ~ ∞), the rotor
tection configuration. earth-current after charging of the earth capaci-
tance is nearly zero. In the event of a fault, the
earth resistance – including coupling resistance
(7XR6004) and the incoming supply voltage – de-
termines the steady-state current. Via the second
input (control input), the transfers, the current
square-wave voltage and the polarity reversal
frequency are recorded. Fault resistances up to
80 kΩ can be detected with this measuring
principle.

Monitoring of the rotor earth circuit for interrup- in generators with high earth capacitances – e.g.
tion takes place by evaluating the current during large hydroelectric generators – is enabled. Angle
polarity reversals. faults due to the earthing or star-point trans-
former are detected during commissioning and
Setting instructions (1 to 3 Hz protection) corrected in the algorithm. The protection func-
Since the protection calculates the resistive rotor tion has a warning and trip stage. In addition, the
earth resistance directly from the values of applied measuring circuit is monitored and a failure of the
voltage, series resistance and flowing earth-current,
the limit values for the alarm stage (RE WARN) and 20 Hz generator detected. Regardless of the earth
the trip stage (RE TRIP) can be set immediately as resistance calculation, the protection function ad-
resistance values. In most cases the preset values ditionally evaluates the r.m.s. value of the current.
(RE WARN = 40 kΩ and RE TRIP = 5 kΩ) are suffi- Another stage is available for earth-faults in which
cient. Depending on the insulation resistance and the displacement voltage and thus the fault cur-
coolant, these values can be changed. It is impor- rent exceed a certain value.
tant to pay attention to an adequate margin be-
Taking the following parameters into considera-
tween the setting value and the actual insulation
tion, the following settings for the application
resistance. As a result of possible disturbances due
example apply.
to the exciter device, the setting for the alarm stage
is finally determined during the primary tests. n Load resistance on the earthing transformer
The delay is usually set for the alarm stage (T RE RL = 4.63 Ω
WARN) to about 10 s, and for the trip stage (T RE n Transformation ratio, voltage divider
TRIP) to a short time of about 1 s. üKl = 200 / 5
The set times are additional time delays which do n Transformation ratio, voltage divider
not include the operating times (measuring time, üdivider = 2 / 5
dropout time) of the protection function. n Transformation ratio, earthing transformer
ütransf = 15.75: 3 / 0.5 kV
Parameter Setting options Default *)
Parameter Setting Setting
Pickup value of the 5 to 80 kΩ 40 kΩ
options
alarm stage
Pickup value of the 20 to 700 Ω 193 Ω
Pickup value of the 1 to 10 kΩ 5 kΩ
alarm stage SES 100 %
trip stage
Pickup value of the 20 to 700 Ω 48 Ω
Delay time of the 0 to 60 s; ∞ 10 s
trip stage SES 100 %
alarm stage
Delay time of the alarm stage 0 to 60 s; ∞ 10 s
Delay time of the 0 to 60 s; ∞ 1s
SES 100 %
trip stage
Table 13 Parameter overview for the rotor earth-fault Delay time of the trip stage 0 to 60 s; ∞ 1s
protection SES 100 %
Pickup value 100 % I>> 0.02 to 1.5 A 0.27 A
Monitoring threshold for 0.3 to 15 V 1V
4.12 100 % stator earth-fault protection with 20 Hz voltage
20 Hz injection (ANSI 64 G (100 %))
Monitoring threshold for 5 to 40 mA 10 mA
The injection of a 20 Hz voltage for detection of 20 Hz current
faults in the star point or close to the star point of
Angle correction for I SES 60 ° 0°
generators has proven a safe and reliable method.
Unlike the 3rd harmonic criterion (see page 12, Transition resistance Rps 0 to 700 Ω 0Ω
Catalog SIP 6.1), it is independent of the genera- Parallel load resistance 20 to 700 Ω; ∞ ∞Ω
tor properties and the operating method. Mea- Table 14 Parameter overview for the 100 % stator earth-fault
surement at system standstill is still possible. This protection
protection function is designed so that it detects
earth-faults both in the whole generator (real
100 %) and in all galvanically connected system
components. The protection relay detects the in-
jected 20 Hz voltage and the flowing 20 Hz cur-
rent. Disturbance variables such as stator earth
capacitances are eliminated, and the ohmic fault
resistance is determined by a mathematical model.
As a result, high sensitivity is ensured and the use

4.13 Out-of-step protection (ANSI 78)

This protection function serves to detect power
swings in the system. If generators feed too long
onto a system short-circuit, a compensation pro-
cess (active power swings) may take place be-
tween the system and the generator after fault dis-
connection. If the center of power swings is in the
area of the unit, the “active power surges” lead to
impermissible mechanical stressing of the gener-
ator and the whole generator mounting – includ-
ing the turbine. Since these are symmetrical proc-
esses, the positive-sequence impedance is calculat-
ed from the voltage and current positive-sequence
components and the impedance curve is evalua-
ted. In addition, the symmetry is monitored by
evaluating the negative-phase sequence system
current. Two characteristics in the R/X diagram
describe the range of effect (generator, unit trans-
former or system) of the out-of-step protection.
The appropriate counters are incremented, de-
pending on in which characteristic range the im-
pedance vector enters and exits. If the set counter Fig. 7 Power swing polygon
reading is reached, tripping takes place. If no more
power swings occur after a set time, the counters
The meaning, calculation and setting of the pa-
are automatically reset. Every power swing can be
rameters for the trip characteristics are described
signalled by a settable pulse. The extending of the
in detail in the manual.
characteristic in R direction determines the de-
tectable power swing angle. 120 ° are practicable. The following table shows the setting options and
The characteristic can be tilted at an adjustable the calculated settings.
angle to adapt to the conditions when the system
is feeding off several parallel generators. Parameter Setting options Setting
Pickup value of the 20 to 400 % 120 %
Setting instructions measurement release I1>
A minimum value of the positive-sequence com-
Pickup value of the 5 to 100 % 20 %
ponents of the currents I1> must be exceeded measurement release I2>
(overcurrent pickup) to enable the measurement.
Resistance Za of the polygon 0.2 to 130 Ω 8.25 Ω
In addition, a maximum value of the negative- (width)
sequence components of the currents I2< may not
be exceeded, due to the symmetry condition. As a Reactance Zb of the polygon 0.1 to 130 Ω 19.60 Ω
(reverse)
rule, the setting value I1> is chosen above rated
current – i.e. about 120 % IN – to avoid pickup by Reactance Zc of the polygon 0.1 to 130 Ω 8.90 Ω
(forward char. 1)
overload. The pickup threshold of the negative-
sequence component of the current I2< is set to Reactance difference 0 to 130 Ω 1.10 Ω
char. 2 – char. 1
about 20 % IN.
Inclination angle of the polygon 60 to 90 ° 90 °
The impedances of the protected zone seen from
the protection relay are decisive for determining Number of oscillations by 1 to 4 1
characteristic 1
the setting values. In the direction of the generator
(seen from the installation position of the voltage Number of oscillations by 1 to 8 4
characteristic 2
transformer set), the power swing reactance of the
generator must be taken into account; it can be set Seal-in time of characteristic 1 0.2 to 60 s 20 s
approximately equal to the transient reactance xd'. and characteristic 2
This means the transient reactance related to the Seal-in time of the message 0.02 to 0.15 s 0.05 s
secondary side is calculated and set for Zb xd' (see out-of-step, char. 1 and
out-of-step char. 2
Fig. 7).
Table 15 Parameter overview for out-of-step protection
The setting ranges and presettings are specified for

a secondary rated current of IN = 1 A.

n 5. Connection diagram n 6. Communication

The 7UM6 relays fully meet the requirements of
modern communication technology. They have
interfaces which allow integration in master con-
trol stations, convenient parameterization and
operation by PC (locally or via a modem). The
7UM6 supports the widely used international
open communication standards
n PROFIBUS-DP, RS485 or optical 820 nm
double-ring ST connector
n IEC 60870-5-103,
n DNP3.0, RS485 or optical 820 nm double-ring
ST connector and
n MODBUS, RS485 or optical 820 nm
double-ring ST connector
Note
All SIPROTEC 4 relays also operate with star
coupler. This enables the user to access all infor-
mation from the office or en route (for simple ap-
plications). With the PROFIBUS-DP protocol,
SIPROTEC relays can easily be integrated in PLC-
based process control systems (e.g. SIMATIC
S5/S7). The protocols DNP3.0 and MODBUS
ASCII/RTU allow integration in numerous instru-
mentation and control systems of other manu-
facturers.
n 7. Summary
Beginning with the recommendations for protec-
tion functions [1], it has been described that effi-
cient concepts can be created with modern
SIPROTEC relays in medium-sized generators,
despite the need to consider cost factors. The mul-
tifunctional, numerical SIPROTEC relays enable a
greater functional scope than the previous single
relays. Self-monitoring substantially improves the
availability of the protection relays.
For further information about the function range
and setting, the 7UM62 manual is recommended,
Fig. 8 Connection diagram of 7UM6 chapter 2 of which has been compiled as an appli-
cation manual.
n 8. References
Herrmann, H.-J.: “Digitale Schutztechnik”
(Digital Protection Technology).
Basic principles, software, examples of
implementation.
VDE-Verlag GmbH, Berlin 1997,
ISBN 3-8007-1850-2.
Herrmann, H.-J.: “Elektrischer Schutz von
Kleinkraftwerken” (Electrical Protection of Small
Power Stations).“Elektrizitätswirtschaft Jg. 97”
(Electricity Industry Year 97) (1998) Issue 24
Manual
7UM62 Multifunction Generator, Motor and
Transformer Protection Relay

Unit Protection System for

Pumped-Storage Power
Stations
n 1. Introduction
In many power systems, pumped-storage power
stations are used in addition to run-of-river power
stations. These power stations serve primarily to
cover load peaks in the power system. In such
peak load periods the machines operate as genera-
tors and feed active power into the grid. In slack
periods, e.g. during the night, the machines oper-
ate as motors and pump water into the upper res-
LSP2171.eps
ervoir which is then available later as an energy
source for peak load supply. In this way, large-
scale thermal power units can be continuously run
to cover basic load.
Fig. 2 shows a typical redundant protection system
design for pumped-storage units with an active
Fig. 1 SIPROTEC 7UM62
power output greater than 100 MW.
Fig. 2 Redundant protection system for pumped-storage power stations

If the protection requirements of a pumped-stor- n 2. Selected protection functions

age power station are compared with those of a 2.1 Stator earth-fault protection
unit for pure energy generation, certain particular The same principle applies for generators in
features emerge as well as many parallels. Design- pumped-storage power stations as for other power
ing a protection system for medium and large- station types: the most frequent electrical fault is
scale power stations is described in another publi- an earth fault. Simple stator earth-faults cause no
cation. This guide explains the special protection damage to the generator provided the earthing
functions and the working of a pumped-storage transformer and loading resistor have been cor-
unit. rectly designed. A second earth fault would, how-
Table 1 below describes a typical basic concept for ever, result in such high fault currents that the
a pumped-storage unit. generator would be severely damaged. For this
reason a simple stator earth-fault must be detected
Protection functions ANSI code Protection Group A Protection Group B at an early stage and at least signalled. In large
generators simple earth faults nearly always result
Stator earth-fault protection 90 % 64
in disconnection unless restrictions in the power
Stator earth-fault protection 64 (100 %) n n system operation and control exclude this.
100 %
Differential protection 87 n
The 100 % stator earth-fault with 20 Hz infeed
(Fig. 3) complies with the requirements for reli-
Winding fault n able detection of a stator earth-fault throughout
Overcurrent-time protection 51 n the stator winding up to the star point. With
Impedance protection 21 n infeed of a generator-independent auxiliary volt-
age, this protection principle allows so-called
Rotor earth-fault protection 64R n n standstill testing. External supply to the 20 Hz
Load unbalance protection 46 n n generator enables the machine to be tested for
Underexcitation protection 40 n possible stator earth-faults before startup. In nor-
mal operation the sensitive signal level indicates
Out-of-step protection 78 n
the start of an earth fault early on. The protection
Stator overload protection 49 n n relay calculates the resistive component from the
Rotor overload protection 49E n measured earth current. The result is therefore in-
dependent of the capacitance of the stator to
Overvoltage protection 59 n n
earth. This advantage is especially effective in large
Frequency protection f > 81 (f >) n n hydroelectric generators which have a high capaci-
Frequency protection f < 81 (f <) n n tance to earth. The insulation value of the earthed
stator winding can be shown on the protection
Reverse power protection 32 n
relay display, providing the operator with valuable
Underpower protection 37 n information about whether the machine can con-
Undervoltage protection 27 n tinue to run until the next planned inspection
shutdown.
Overexcitation protection 24 n
Table 1 Protection concept for pumped-storage power plants 2.2 Negative-sequence protection (ANSI 46)
Various settings are required for phase-se-
The protection system described can be supple- quence-dependent protection functions if the ma-
mented by extra protection functions, depending chine is switched between generator operation
on technical requirements or country-specific cus- and pump operation. An example of such a func-
tomary protection concepts. Likewise, the level of tion is negative-sequence (or unbalanced load)
function redundancy can be expanded or reduced. protection, whose pickup criterion is based on
The advantage of the 7UM6 multifunctional pro- measuring or calculating the reverse field. One
tection relays lies in the ease of configuration. possible solution is installing two identical protec-
tion relays with respective setting values for gener-
With the aid of the DIGSI® configuration pro - ator operation and pump operation. One of the
gram, each of the protection functions contained two protection relay is always blocked to prevent
in the relay can be switched active or inactive. spurious operation.

A more elegant and cost-effevtive solution would

be to switch over the parameter sets in the protec-
tion relay when changing between the two operat-
ing modes. The 7UM6 numerical generator pro-
tection relay allows the read-in of two independ-
ent parameter sets. When changing between the
machine's two operating modes, a control signal
to the protection relay digital input activates the
other parameter set for protection processing. For
negative-sequence protection, this means that,
from the calculated symmetrical components of
the measured current, the positive phase-sequence
system and the negative phase-sequence system
are exchanged for further protection processing.

Large pumped-storage units are often used for re-
active power compensation, in addition to balanc-
ing peak loads in the power system. Consequently,
the generator can be run close to its stability limit,
depending on reactive-power demand over a long
period. Slight increases in reactive-power demand
can make the generator fall out of step. Conse- Fig. 3 Stator earth-fault protection with generator, 20 Hz
quently, particular emphasis is laid on underexci-
tation protection in large pumped-storage power
stations.
Primarily, the excitation system itself detects an
overshoot of the stability characteristics of the
generator and compensates this by readjusting the
excitation voltage. Electrical underexcitation pro-
tection comes into action if this automatic read-
justment is unsuccessful. The generator moves
into the capacitive working range and finally falls
out of step. This impermissible operating state can
damage the generator and create instability in the
power system linked to the power station. The
asynchronously running generator suffers severe
mechanical stressing of the shaft and bearings,
which could result in costly repair work. It also
produces electrical oscillations in the power sys-
tem which can lead to other generators falling out Fig. 4 Characteristic of underexcitation protection
of step. Asynchronous operation can be caused by
defective excitation equipment or excessive reac- These protection characteristics form the stability
tive-power demand in the connected transmission characteristics of the machine and combine high
network. availability of the pumped-storage unit with
optimum protection of the machine and the con-
In view of the wide range for active and reactive nected power system.
power in such a pumped-storage unit, particular
attention must be paid to setting the underexcita- The pickup characteristics of the underexcitation
tion protection characteristic. Simple detection of protection are made up of the machine’s steady-
reactive-power demand by a circular protection state and dynamic stability characteristics. An
characteristic will not meet these operating re- overshoot of the steady-state characteristic is an
quirements. A combination of pickup characteris- initial indication that stability is endangered.
tics in accordance with Fig. 4 is ideally suited for
the requirements described here.

If the excitation equipment is functioning cor- This logical linking of stator criteria with rotor
rectly the machine can be returned to synchro- criteria enables the pumped-storage unit to con-
nous operation again by automatically readjusting tinue operation up to its real stability limit. The
the excitation voltage. Only an overshoot of the exact replica of the machine characteristics en-
dynamic restraint characteristic finally makes the sures that the unit does not switch off unnecessar-
machine fall out of step. Both characteristics, ily within the permissible limit range. On the
sometimes referred to as the stator criterion, are other hand, this optimum utilization of the oper-
calculated from machine currents and voltages. ating range involves no danger to the machine or
The excitation voltage Uexcit (or rotor criterion) is power system. If the stability of the machine is no
used as an additional measured variable for opti- longer guaranteed, it is shut down quickly and
mum operation and control. The recommended safely.
function logic of the underexcitation protection is
shown in Table 2 below. 2.4 Stator overload protection
Measurement criterion Protection reaction An important protection function for pump oper-
ation is stator overload protection. In generator
Steady-state characteristic Indication (optional
tripping approx. 10 s)
mode, the thermal loading on the stator winding
is limited by the turbine output if the power sta-
Steady-state characteristic Tripping approx. 0.3 s tion unit is correctly designed. In pump mode,
+ Uexcit <
there may be thermal overload of the machine in
Dynamic characteristics Tripping approx. 0.3 s motor mode. A thermal replica with complete me-
Uexcit < Indication mory prevents such overloads. Particularly with
Table 2
quick load changes, the thermal replica is more
exact than a direct temperature measurement on
An overshoot of the steady-state stability charac- the winding.
teristic without simultaneous excitation voltage The thermal replica calculates the temperature of
dip is only signalled. The operating personnel can the stator winding from the stator current practi-
manually intervene to return the machine to sta- cally instantaneously (Fig. 5) according to the
ble operation. The cause of underexcitation can lie formula I2t. A direct measurement of the tempera-
in power factor correction. ture on the insulation of the stator winding fol-
Only a simultaneous drop in the excitation voltage lows load changes only after a delay and therefore
below a set value causes shutdown of the machine does not always show the current temperature of
in short time. This is probably due to a fault in the the conductor.
excitation equipment. It is then only a question of
time before the machine reaches the dynamic sta- 2.5 Rotor monitoring
The setting value IB is matched
to the rated motor current bility characteristic and finally falls out of step. In large machine units special attention must be
The rated temperature ΘN is Shutting down at an early stage prevents this addi- paid to monitoring the rotor circuit, owing to the
reached at I = IB high power per pole. This monitoring includes in-
tional pole slipping and protects both machine
The tripping temperature ΘA is
reached at I = 1.1 IB and power system. sulation of the rotor winding as well as the ther-
mal load on the rotor winding.
To monitor the rotor insulation a specially devel-
oped measurement principle is used which largely
compensates for the disturbing influence of the
rotor capacitance to earth and fluctuations of the
excitation voltage. A reverse polarity square-wave
voltage in a clock pulse from 1 to 3 Hz is applied
between rotor circuit and earth (see Fig. 6). The
steady-state circulating current remaining after
the capacitive charging current has decayed is a
measure of the insulation resistance of the rotor
winding. The protection relay measures only the
resistive component of the earth impedance
regardless of the level of earth capacitance.
Fig. 5 Principle of overload protection

Fig. 6 Principle of rotor earth-fault protection
The differential measurement of the resistive earth plica with complete memory and enables flexible
current from positive and negative polarity com- operation and control of the unit within the per-
pensates on the one hand the disturbance of the missible limits, with simultaneous safe protection
excitation voltage and also stops measurement er- against overloads.
rors owing to operation-related changes in the ex-
citation voltage. This sophisticated measuring 2.6 Power system failure in pump operation
procedure measures an insulation resistance The power-shortfall protection function is only
(to earth) in the rotor winding up to the order of activated in pump mode. This happens by way of
80 kΩ. Hence, the protection relay detects a rotor the above-described parameter set changeover
earth-fault as it arises. The operator can plan the when the mode is changed. In the parameter set
necessary maintenance work on the machine for for generator mode, the power-shortfall function
the long term. Protection-related shutdown of the is deactivated. This protection function recognises
pumped-storage unit is thereby almost always the sudden failure of the power system supply
prevented which considerably enhances the avail- (when the machine is in pump mode) by measur-
ability of the unit. ing the active power in the infeed direction. If a
As a result of the various operating modes of the fault is detected, the pumped-storage unit is
pumped-storage unit switched off and the spherical valves are closed.
n Generator mode A second measurement principle for detecting a

line side supply failure is underfrequency protec-
n Generator and VAR compensation mode
tion, which is likewise included in the protection
n Pump mode concept.
n Pump and VAR compensation mode
n Any electrical brake The underfrequency protection is not in operation
during run-up and switchover of the system’s var-
the rotor circuit of the excitation system may be ious operating states, and is only activated after
exposed to high thermal loads. Overload protec- the machine has been synchronized with the
tion on the high-voltage side of the excitation power system. A binary input from the protection
transformer measures the current flowing there relay, which requests the position of the system
which is in direct relationship to the thermal rotor circuit-breaker, effects this control.
load. Overload protection works as a thermal re-

n 3. Summary
In view of their particular status in hydropower
engineering, pumped-storage power stations place
special demands on the electrical protection sys-
tem. Owing to the various operating modes
n Generator mode for energy supply
n Pump mode to feed back energy
n VAR compensation mode
a protection system is required that automatically
adapts itself to these changing operating states.
The SIPROTEC 7UM6 relays are specifically de-
signed to handle these variable operating condi-
tions. This applies both to the protection
functions comprised in the relays and to the flexi-
ble adaptation of the protection system via exter-
nal control signals from the power station.

Busbar Protection
Application of Low-Impedance
7SS601 Busbar
Differential Protection
n 1. Introduction
Utilities have to supply power to their customers
with highest reliability and minimum down time.
System disturbances, especially short-circuits,
cannot always be avoided. They are caused by hu-
man error, accidents, nature’s influence such as
storm, lightning, etc...
However, damage to primary equipment, such as
LSP2363.eps
transformers, switchgear, overhead lines, etc. must
be limited in order to reduce the repair time and,
thus, the downtime.
Although busbar faults are rare, they are consid-
ered most dangerous for people (staff) and the
switchgear. Hence, fast tripping in case of busbar
faults is essential!
This can be achieved primarily by differential pro- Fig. 1 SIPROTEC 7SS601 centralized numerical busbar
tection. protection relay
Especially in medium and high-voltage switch- n 2. Principle of low-impedance protection

gear, busbar differential protection is often re-
Fig. 2 illustrates the basic principle of the low-
garded as an expensive accessory. However, re-
impedance protection.
sponsible engineers are aware of the risk of
extensive outages, if busbar faults are not cleared
fast and selectively.
Siemens offers with its 7SS601 low-impedance
busbar differential protection a low-cost solution,
particularly suitable for single busbar with or
without bus sectionalizer or simple double busbar
configurations.
Fig. 2 Principle of differential protection
The 7SS601 combines important advantages of
numerical protection with a cost-effective and an
easy-to-use protection system: Differential protection is based on the Law of
Kirchhoff: in a healthy system, the sum of currents
n Self-supervision, fault logging and event record- in a node must be zero. This is the ideal case. But
ing, setting with DIGSI required for only a few CT errors and measuring errors need to be con-
parameters. sidered. For that reason, the protection needs to
n Highest flexibility with regard to busbar config- be stabilized.
uration, number of feeders, different CT-ratios, The differential criteria and the restraint criteria
low CT requirements. are defined as follows:
n Measured-value acquisition by summation or Differential current: IDiff = |I1 + I2 + …. In |
matching current transformers (phase-selective) Restraint current: IRestraint = |I1| + |I2 |+ ….| In|
n Fast and selective tripping for all busbar faults.
n Suitable for all voltage levels, up to 500 kV Fig. 3 shows, how IDiff and IRestraint are being derived.
The following article describes the basics of low- It can be seen, that for load or through-flowing
impedance busbar protection; some of its typical currents the differential criteria is almost zero,
applications on a single busbar with bus whereas the restraint quantity rises instantly.
sectionalizer with disconnect switch or bus In case of an internal fault both, the differential
sectionalizer with circuit-breaker). and the restraint quantity rise at the same time.
The example is tailored for a solidly earthed net- Hence, even within a few milliseconds, the protec-
work. tion relay can decide whether there is an internal
or external fault.

Busbar Protection
Fig. 3
Derivation of
IDiff and IRestraint For more than 50 years Siemens has been using this n 3. Protected objects
kind of stabilized differential protection. Fig. 4 and 5 show examples with 10 feeders on
It was introduced for the first time in the electro- each bus section. However, the number of feeders
mechanical protection 7SS84, later in the analog per bus section is not limited.
static protection 7SS10 and is now being employ-
ed in the numerical protections 7SS52 and 7SS601. 3.1 Single busbar with disconnector (DS) in bus
sectionalizer. In this case, no CTs are used in the
bus sectionalizer. See Fig. 4
3.1.1 If the DS is open, both measuring systems
(bus differential schemes) work independently. In
case of a busbar fault, only the circuit-breakers
(52) connected to the faulty bus will be tripped.
3.1.2 If the DS is closed the busbar must be con-
sidered as “one unit”. The preference circuit (pref -
erential treatment module) will switch the cur-
rents of all feeders to one measuring system only.
In case of a busbar fault, all circuit-breakers will
be tripped.
3.2 Single busbar with circuit-breaker and a CT in
Fig. 4 Single busbar with disconnector in bus
sectionalizer the bus sectionalizer. See Fig. 5.
3.2.1 If the circuit-breaker is open, both systems
work independently. In case of a busbar fault, only
the CBs connected to the faulty bus will be trip-
ped.
Since the circuit-breaker is open, the CT is not re-
quired for measurement and thus shorted.
In case of a fault between the circuit-breaker and
the CT in the bus sectionalizer, system 1 will de-
tect the fault as “internal” and trip all CBs con -
nected to bus 1.
Fig. 5 Single busbar with circuit-breaker in bus

sectionalizer

Busbar Protection
3.2.2 If the circuit-breaker in the bus sectionalizer

is closed both systems work independently. In case
of a busbar fault, all circuit-breakers connected to
the faulty busbar will be tripped. The circuit-
breaker in the bus sectionalizer will be tripped by
both systems. In case of a fault between the cir-
cuit-breaker in the bus sectionalizer and the CT,
system 2 will detect this fault as “internal” and trip
all circuit-breakers of bus 2 including the circuit-
breaker in the bus sectionalizer.
System 1 remains stable for the time being.
Once the circuit-breaker in the bus sectionalizer has
tripped, the CTs are shorted. Now system 1 will de-
tect this fault also as “internal” and finally clears the
fault by tripping the circuit-breakers of bus 1.
n 4. Summation current transformers and

adaptation of different CT ratios
The advantage of the low-impedance scheme is
that other protection relays may be connected in Fig. 6 Serial connection of summation CTs with other relays
series with the summation or matching current
transformers (Fig. 6). This reduces the overall
costs of the switchgear.
As mentioned earlier, the low-impedance scheme
can process different CT ratios. Thus, upgrading
of existing switchgear with low-impedance busbar
protection is easily possible without changing or
adding CT cores!
The standard connection of a summation current
transformer is shown in Fig. 7.
The summation current transformer is used to do a
“magnetic” summation of the currents. The result-
ing currents depend on the ratio of the windings.
The following calculation shows some examples of current
how the currents of the summation current trans-
formers are being calculated. Fig. 7 Standard connection of summation current transformer
It can be shown that the optimum ratio of the pri- WP = primary winding of summation current transformer
Ws = secondary winding of summation current transformer
mary windings are 2:1:3. This ensures increased
sensitivity for earth faults.
Example: WP1 = 60 windings WP2 + WP3 120
i L3 = 1 A =1 A = 0.24 A
WP2 = 30 windings WS 500
WP3 = 90 windings
ß 0.24 ⋅e j240 ß iL3 = -0.12 - j0.21
WS = 500 windings ( fixed )
General equation: Result: summation iL1 + iL2 + iL3 = is
W
i p ⋅ WP = i s ⋅ WS ⇒ i s = i p P
WS
. ⋅ e − j 30 °
is = 0.09 - j0.054 ß i s = 0105
For the above example: assume i = I N = 1 A
WP1 + WP3 150

i L1 =1 A =1 A = 0.3 A ß 0.3 ⋅e j0
WS 500
ß iL1 = 0.3 + j0
WP3 90
i L2 = 1 A =1 A = 018 . ⋅e j120
. A ß 018
WS 500
ß iL2 = -0.09 + j0.156

Busbar Protection
Wind- Refer- Phase Connec- Links

Above example
ings ence tions
6-3-9 1 L1 C, D
L3 A, B
N E, F
12-6-18 2 L1 A, F B-E
L3 C, D
N G, H
18-9-27 3 L1 G, H
L3 E, F
N A, K B-J
24-12-36 4 L1 J, K
L3 A, F B-E
N L, M
Fig. 8 Phasor diagram of measurings 30-15-45 5 L1 C, K D-J

L3 B, H A-G magnetic
subtraction
As can be seen the secondary current is N E, M F-L
≈ 100 mA, which corresponds to the rated current 36-18-54 6 L1 A, K B-E, F-J
of the 7SS601 protection relay. L3 G, H
The graphical addition leads to the same result. N M, O L-N magnetic
subtraction
See Fig. 8.
42-21-63 7 L1 C, M D-L
In order to adapt different CT-ratios, the L3 B, K A-J magnetic
4AM5120 summation current transformer has 7 subtraction
primary windings, which can be combined by N F, O E-H, G-N magnetic
subtraction
means of connecting the windings in series.
The 4AM 5120-3DA00-0AN2 summation current 48-24-72 8 L1 A, M B-E; F-L
transformer is suitable for 1A rated current: L3 J, K
N H, O G-N magnetic
subtraction
54-27-81 9 L1 G, M H-L
L3 A, K B-J
N F, O E-N magnetic
subtraction
60-30-90 10 L1 J, M K-L
L3 A, H B-E, F-G
Fig. 9 Tapping of summation current transformers N N, O
As mentioned before, the ratio of the primary Table 1 Selected ratios of summation current transformers
windings should be 2:1:3 in standard applications. at IN = 1 A
Table 1 shows the most commonly used ratios.
n Choose the smallest common integer multiple
Note
of the CT-ratios, of which the result of division
Always be aware of the orientation of the windings !
may not exceed “10”.
The examples of Figures 4 and 5 respectively show Example
CT-ratios of 400/1A, 600/1A and 1000/1A. 400 / 600 / 1000 smallest multiple:
The differential protection can only compare cur- 2 → 200 / 300 / 500: Not possible !
rents, if the basis for comparison is equal, i.e. the 400 / 600 / 1000 smallest multiple with result
CT-ratios must be matched. ≤ 10 : 100 → 4 / 6 / 10. Possible !
A simple procedure has to be followed:
The result of this calculation is used to select the
n The difference in CT-ratios may not exceed 1:10 ratios of the summation transformer from Table 1
(i.e. 400 / 600 / 1000 is possible, 200 / 800 / 2500 ( Reference number ).
is not possible).
400 / 1A → 4 → 24-12-36
n The max. possible number of windings of the
600 / 1A → 6 → 36-18-54
summation current transformer shall be used.
1000 / 1A → 10 → 60-30-90
Thus, accuracy is increased.
n The highest CT-ratio is always the reference.

Busbar Protection
Fig. 10 shows that with the above selected ratios

the secondary currents of the summation current
transformers are equal. The differential protection
can now compare all currents.
In case of short-circuits, the sensitivity of the dif-
ferential protection varies due to the winding ratio
2:1:3 and the resulting secondary currents. Refer
to Table 2.
Short Effective W I1 for iM = 100 mA
circuits windings 3
W
L1 - L2 - L3 3 1.00 1.00 ⋅ IN
L1 - L2 2 1.15 0.87 ⋅ IN
L2 - L3 1 0.58 1.73 ⋅ IN
L3 - L1 1 0.58 1.73 ⋅ IN
L1 - E 5 2.89 0.35 ⋅ IN
L2 - E 3 1.73 0.58 ⋅ IN
L3 - E 4 2.31 0.43 ⋅ IN
Table 2 Result of secondary current matching
To ensure reliable tripping, the minimum short-

circuit currents must be above the lowest pickup
value of the respective fault type.
Example: Setting of the pickup threshold:
IDiff > 1.20 INO*
* INO is the rated current of the reference ratio (1000/1 A)
1000 / 1 A
3-phase faults 1.20
Fig. 10 Adaption of different CT ratios
L1 - L2 1.04
L2 - L3 2.08
L3 - L1 2.08
L1 - E 0.42
L2 - E 0.69
L3 - E 0.52
Table 3 Sensitivity of 7SS60 according to type of fault
Hence, the minimum short-circuit current must

exceed:
1200 A for 3-phase faults
2080 A for phase to phase faults
690 A for phase to ground faults

Busbar Protection
n 5. Components of the 7SS601 centralized n 7TR71 preferential treatment/isolator replica

numerical busbar protection module. For allocation of current and preferen-
The basic principles are shown in Fig. 11 and 12: tial treatment via isolator replica.
4AM5120 summation current transformers, re- n 7XP20 housing to accommodate 7TM70 and
straint/command output 7TM70 modules, 7TR71 7TR71 modules. One housing can accommo-
isolator replica/preferential treatment module date up to 4 modules.
and 7SS601 protection relay are required: Fig. 11 shows the scheme with a disconnector in
n Summation current transformers: one for each the bus sectionalizer. In this case, 7TR71 is used as
feeder. preferential treatment module. If the disconnector
is closed the entire bus must be seen as “one unit”.
n Restraint modules: each module 7TM70 con-
All currents must be measured by one system on-
tains 5 input transformers with rectifiers and
ly. Thus all currents are routed to system 2.
5 tripping relays.
The trip circuits of both busses are switched in
n Measuring system(protection relay): one for parallel.
each bus section. Fig. 11 corresponds to the example shown in Fig. 4.
Fig. 11 Block diagram of the 7SS60 with disconnector in bus sectionalizer

Busbar Protection
Fig. 12 shows the scheme with a circuit-breaker in Fig. 12 corresponds to the example shown in
the bus sectionalizer. In this case, 7TR71 is used as Fig. 5.
“circuit-breaker position replica” module. If the
circuit-breaker is open, the secondary side of the
summation CT is shorted because no current flows
in the bus section anyway.
If the circuit-breaker is closed, the differential and
restraint currents are fed to both measuring sys-
tems with opposite direction.
In case of a busbar fault, a busbar-selective tripping
is possible. The circuit-breaker in the bus sectiona-
lizer will be tripped by both measuring systems.
Fig. 12 Block diagram of the 7SS60 with circuit-breaker in bus sectionalizer

Busbar Protection
Busbar protection for switchgear (20 feeders) as Fig. 13 shows the tripping characteristic of the
shown in Fig. 11, comprises the following com- protection relay.
ponents:
The threshold Id > should be set above max. load
n 20 x 4 AM 5120 summation current transfor- current (e.g. 1.2 · ILoad) to avoid tripping by the
mers (10 feeders on each busbar) load current in case of a fault in the CT circuit.
n 4 x 7TM70 restraint modules If, however, the minimum short-circuit currents
(4 x 5 inputs = 20 inputs) require a lower setting, additional trip criteria may
n 2 x 7SS601 protection relays be introduced (e.g. voltage).
n 1 x 7TR71 preferential treatment module On the other hand, to ensure tripping under mini-
n 2 x 7XP20 housings mum short-circuit conditions, Id> should be set at
about 50 % below minimum short-circuit cur-
Busbar protection for switchgear (20 feeders + bus
rents. For instance: Iscmin = 3000 A → 50 % =
sectionalizer with circuit-breaker)
1500 A
n 21 x 4AM5120 summation current transfor- Id > = 1.2 INO, if the reference ratio is 1000/1 A.
mers (20 feeders +1 for the sectionalizer)
The threshold Id>CTS is the pickup value for CT
n 5 x 7TM70 restraint modules (25 inputs) supervision.
n 2 x 7SS601 protection relays If a CT secondary circuit is open or shorted, a dif-
n 1 x 7TR71 isolator/CB replica module ferential current will appear. The differential
n 2 x 7XP20 housings protection will be blocked and an alarm given.
This will avoid unnecessary overfunction in case
The above modules can be accommodated in a
of heavy through-flowing currents.
standard protection cubicle.
Please refer to our documentation (instruction The k-factor changes the slope of the tripping
manual, catalog and circuit diagrams) for more characteristic as shown in Fig. 13 and thus deter-
details. mines the stability of the protection.
Although a high setting for this factor improves
n 6. Setting and design considerations the stability with regard to faults outside the pro-
CTs shall be dimensioned such that all CTs trans- tected zone, it reduces the sensivity for the detec-
form currents without saturation for at least tion of busbar faults. The k-factor should
≥ 4ms. therefore be chosen as low as possible and as high
The number of feeders connected in parallel to as necessary. If the measuring system (protection
one protection relay is unlimited. relay) is to be used for zone-selective protection,
which will be the case in most applications, it is
Please refer to the instruction manual in case of advisable to use the presetting of 0.6 of the k-fac-
systems with transformers of which the star point tor.
is isolated.
A lock-out function of the trip command may be
activated in the 7SS601. No external lock-out re-
lays are required !
Fig. 13 Tripping characteristic

Busbar Protection
n 7. Comparison of high and low-impedance

busbar protection
Nowadays high impedance protection is still
widely used, because it is considered “cheap and
easy”.
But most users only look at the relay price itself,
without considering the additional costs for the
switchgear and other disadvantages of a high-im-
pedance schemes:
n All CTs must have the same ratio
n Class X for all CT cores
n Bus sectionalizers with circuit-breaker must be
equipped with two CTs
n Separate CT cores for busbar protection
n Advantages of numerical protection technology
(e.g. fault recording, communication, etc.) not
available
n Check zone needs separate CT cores
n Isolator replica requires switching of CT sec-
ondaries. Additional check zone obligatory.
n 8. Summary
The low-impedance 7SS601 busbar protection is a
cost-effective solution for medium and high-volt-
age switchgear.
Apart from the application described the 7SS601
can also be applied
n With phase-segregated measurement
n On switchgear with double busbars.
For a quotation, we would need the following in-
formation:
n Single-line diagram, showing
– Busbar configuration
– Number of feeders, bus sectionalizers with
disconnector/circuit-breaker
– Ratio of CTs
– Phase-segregated or single phase measurement
– Complete cubicles or components only
For more information please contact your local
Siemens partner.
More details may be obtained from our docu-
mentation (instruction manuals, relay catalog
SIP2004, CD’s). Circuit diagrams for standard
applications are available on the Internet at:
www.siprotec.com

Busbar Protection

Busbar Protection
Basic Busbar Protection by

Reverse Interlocking
Busbars have a particular key role in power trans-

mission and distribution. They are the central dis-
tribution point for many feeders. In the event of a
fault, the short-circuit current on the busbar is
very high, resulting potentially in mechanical de-
struction and the consequent long repair times,
which would affect all feeders. On the high and
extra-high-voltage level, a fast 7SSx busbar pro-
tection relay is used, which, with a tripping time
of <12 ms, limits the damage by busbar faults.
Fast busbar protection is also used in all impor-
tant medium-voltage switchgear.
For basic medium-voltage switchgear with one in-
coming feeder, no special busbar protection is
used (for reasons of economy). In such cases
busbar protection is provided by the time-over-
current relay of the incoming feeder. As shown in
Fig. 1, tripping of the time-overcurrent protection
for the E1 feeder occurs with a grading time of
Fig. 1 Single busbar with feeder protection
300 ms more than the longest grading time of the
A1-A3 feeder protection. The times selected in
Fig. 1 have been taken as examples. The E1 protec-
tion serves as backup protection for each A1-A3
feeder protection.
A busbar fault is however then only disconnected
after 0.9 seconds, which would result in damage of
considerable magnitude.
In the case of single busbars with one defined in-
coming feeder and otherwise only defined outgo-
ing feeders, fast busbar protection can be provided
with no major additional effort by means of re-
verse interlocking. Such busbar configurations are
common in medium-voltage systems and in auxil-
iary supply networks. The time-overcurrent relays
already available for feeder protection are used, as
shown in Fig. 1. An additional benefit is that all
SIPROTEC relays are equipped with at least two
definite-time current stages, which can be blocked
individually.
The reverse interlocking concept is shown in
Fig. 2. With the time-overcurrent protection E1 of Fig. 2 Single busbar with feeder protection and busbar protection by reverse
the incoming feeder, a further stage I>> with a interlocking
time delay of t2 = 50 ms is provided in addition to
stage I> with t1. The expiry of time t2 can be
blocked via the binary input BI1.

Busbar Protection
The tripping threshold of the stages I> and I>> is n Prospects for further applications
set at the same level in accordance with the net- In the case of ring busbars with two incoming
work conditions (approximately 1.5 x Irated). For feeders or single busbars with sectionalizer, a simi-
the time-overcurrent relays of outgoing feeders lar reverse interlocking principle can be applied by
A1-A3, the pick-up signal is allocated to a dedi- way of short-circuit direction detection. This case
cated contact. The pick-up signals of all feeders will be presented in a separate application.
are connected in parallel and given as a blocking
signal to the binary input BI1 of the relay of the
incoming feeder. Wiring is effected by means of
a copper core, looped from panel to panel (see
Fig. 2). This means that a pickup in an outgoing
feeder A1-A3 will block the tripping of the I>>
stage (t2) of the incoming feeder E1.
Function in the case of a fault in the outgoing feeder:

In the case of a fault on the outgoing feeder (see
“F1” in Fig. 2), both stages in relay E1 of the in -
coming feeder pick up. The relay A1 of the out-
going feeder also picks up and issues a blocking
signal via the pick-up signal to binary input BI1 of
the relay E1 of the incoming feeder. The expiry of
t2 is thus blocked. The fault is disconnected from
the relay A1 of the outgoing feeder. The relay E1
of the incoming feeder operates as backup protec-
tion with t1.
Function in the case of a fault on the busbar:
In the case of a busbar fault (see “F2” in Fig. 2),
both stages in relay E1 of the incoming feeder pick
up and t1 and t2 are started. From an A1-A3 out-
going feeder there can be no infeed onto the fault.
Consequently there is no blocking signal. In the
relay E1 of the incoming feeder, time t2 expires
and trips the circuit-breaker after 50 ms. The
busbar fault is thus disconnected within a short
time and the extent of the fault is limited.
n Summary
For busbars with one incoming feeder and radial
outgoing feeders, i.e. without back-feeding, the re-
verse interlocking principle grants an effective and
fast busbar protection. Additional hardware is not
required, because the SIPROTEC devices incorpo-
rate this function in their basic versions. Attention
shall be paid to motor feeders, which may feed a
busbar fault in the generating mode. They cannot
always be treated as feeders without back-feeding.
224 Printed in Germany KGK 01.05 4.0 224 En 101149/6101D6342

Exclusion of liability
We have checked the contents of this manual for
agreement with the hardware and software de-
scribed. Since deviations cannot be precluded en-
tirely, we cannot guarantee that the applications
described will function correctly in any system.
Copyright
Copyright © Siemens AG 2005. All rights reserved.
The reproduction, transmission or use of this
document or its contents is not permitted without
express written authority. Offenders will be liable
for damages. All rights, including rights created by
patent grant or registration of a utility model or
design, are reserved.
Registered trademarks
SIPROTEC, SINAUT, SICAM and DIGSI are
registered trademarks of Siemens AG. The other
names appearing in this manual may be trade
names the use of which by third parties for their
own purposes may infringe the rights of the
owners.
Further application examples can be found on the Internet. Please visit us at: www.siprotec.com
Published by
Siemens Aktiengesellschaft
Power Transmission and Distribution
Energy Automation Division
Postfach 48 06
90026 Nuernberg
Germany
www.siemens.com/ptd Order No.: E50001-K4451-A101-A1-7600

Siemens - Applications For SIPROTEC Protection Relays

Uploaded by

Copyright:

Available Formats

Siemens - Applications For SIPROTEC Protection Relays

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Siemens - Applications For SIPROTEC Protection Relays

Uploaded by

Copyright:

Available Formats

2005

Line Protection in Transmission Systems

n 1. Distance protection with auto-reclosure on

Siemens PTD EA · Applications for SIPROTEC Protection Relays · 2005 3

n 3. Settings in the configuration with DIGSI n 4. Creating the logic flowcharts

Parameterization: (parameter group A, distance

Fig. 4 CFC plan for protection relay A

4 Siemens PTD EA · Applications for SIPROTEC Protection Relays · 2005

7SA6 – protection relay B: 7SA6 – protection relay B:

Fig. 7 CFC plan for protection relay B

6 Siemens PTD EA · Applications for SIPROTEC Protection Relays · 2005

Redundant Supply with

Automatic switchover of incoming supply with

Siemens PTD EA · Applications for SIPROTEC Protection Relays · 2005 7

Switchover with redundant incoming feeders

8 Siemens PTD EA · Applications for SIPROTEC Protection Relays · 2005

In the case of rapid system decoupling (opening

3.1 Practical principle

3.2 Description The relays are interconnected by binary signal

Siemens PTD EA · Applications for SIPROTEC Protection Relays · 2005 9

d) If the circuit-breaker in the faulty incoming

10 Siemens PTD EA · Applications for SIPROTEC Protection Relays · 2005

Siemens PTD EA · Applications for SIPROTEC Protection Relays · 2005 11

2.2 Inverse-time overcurrent protection

The general IEC/BS characteristic is shown in

12 Siemens PTD EA · Applications for SIPROTEC Protection Relays · 2005

Fig. 4 IEC/BS characteristics Fig. 5 ANSI characteristics

Siemens PTD EA · Applications for SIPROTEC Protection Relays · 2005 13

Fig. 6 Example of a 20 kV ring distribution system

14 Siemens PTD EA · Applications for SIPROTEC Protection Relays · 2005

Setting range Setting

Ip_mi = 2.3 kA Ip_ma = 7.8 kA

Overcurrent-time relay with “Normal Inverse” (NI) setting

Siemens PTD EA · Applications for SIPROTEC Protection Relays · 2005 15

Setting range Setting

Ip_mi = 2.3 kA Ip_ma = 7.8 kA

Overcurrent-time relay with “Very Inverse” (VI) setting

16 Siemens PTD EA · Applications for SIPROTEC Protection Relays · 2005

Setting range Setting

Ip_mi = 2.3 kA Ip_ma = 7.8 kA

Overcurrent-time relay with “Extremely Inverse” (EI) setting

Siemens PTD EA · Applications for SIPROTEC Protection Relays · 2005 17

Setting range Setting

Ip_mi = 2.3 kA Ip_ma = 7.8 kA

Overcurrent-time relay with VI setting,

18 Siemens PTD EA · Applications for SIPROTEC Protection Relays · 2005

Setting range Setting

Ip_mi = 2.3 kA Ip_ma = 7.8 kA

Overcurrent-time relay with EI setting,

Siemens PTD EA · Applications for SIPROTEC Protection Relays · 2005 19

For the transformer considered (20/0.4 kV, n Conclusion

20 Siemens PTD EA · Applications for SIPROTEC Protection Relays · 2005

Fig. 2 Block diagram

Siemens PTD EA · Applications for SIPROTEC Protection Relays · 2005 21

2.1.1 Grading and selectivity 2.1.3 Inverse-time overcurrent protection

2.1.2 Definite-time overcurrent protection

Fig. 3 Comparison of inverse-time overcurrent

22 Siemens PTD EA · Applications for SIPROTEC Protection Relays · 2005

2.1.6 Sensitivity n 3. Settings