03 TLP Using Directional Overcurrent Elements Part I r5
03 TLP Using Directional Overcurrent Elements Part I r5
Transmission Line
Protection
Revision Table
Editor Tech Review TDS Review Date Rev # Comments
LGP CLS 1-19-04 1,2 Adapted from “Directional element design and
evaluation”
CRT MF, LGP 1-22-04 3 CRT and LGP corrections
Objectives
z Explain why directional elements are
necessary.
z Describe the connections of traditional
phase and ground directional relays.
z Describe the traditional methods for
directional overcurrent relay application and
coordination.
z Determine graphically and algebraically the
operation of directional elements for a given
fault condition.
IS
Overcurrent Element
IS (50 and/or 51) Trip
CT CB IP Line
IS
Trip
The overcurrent relay receives its input current from the secondary circuit of a
current transformer. The current transformer’s characteristics are set so that the
relay receives an accurate replica of the primary current. This must be
accomplished for a range of fault current magnitudes. If the CT is well designed,
the secondary current “seen” by the relay can be calculated as:
IS = IP/CTR
Once the current overcomes the pickup, the relay order to trip the circuit breaker,
CB, can take place immediately or with some time delay, depending on the relay
design and application.
Ic 51N
Ib
52
Line
To cover all types of short circuits within lines, four relays are commonly used in
the traditional scheme with electromechanical technology. Three of these relays
are set to cover phase faults and one relay to detect ground faults.
As shown in the figure, the CTs connect with the relays so each of the phase fault
relays measures the line currents. The pick-up current of these relays must be set
to a relatively high value, to avoid relay operation during high load conditions.
The residual current in the case of a perfectly balanced system is zero. Thus the
ground relay pickup can be set to relatively low values, to avoid relay operation
due to the existence of some residual current during normal operation conditions.
During ground faults, even with high fault resistances, the zero-sequence current
appears and the ground relay detects the fault. Of course, for extremely high fault
impedance, even the zero-sequence current is close to zero, so the fault might
remain undetectable.
tDT
Ii I IDT I
IIT I IIT Ii I
The operation time of the overcurrent relay depends on the design and application. How the
operating time of the relay changes with the magnitude of the applied current is the primary
consideration. Three types of relay elements can be identified according to their time vs. current
characteristic curve:
Instantaneous overcurrent element operates in a very short time when the applied current
surpasses the threshold. A relay is instantaneous when it is designed with no intentional delay.
Operating times for these relays is from 0.5 to 2 cycles. In these elements only the pickup
current is settable.
Definite time element operates with a constant, or definite time delay for any current over the
pickup current. In these elements the operating time can be set as well as the pickup current.
Inverse time elements have a time-vs-current curve requiring faster operation as the applied
current is larger. These elements are very popular due to their predictable behavior, similar to
the behavior of fuses. The level of the curve can be adjusted by the user.
It is common to find relays which contain several types of elements. The resulting combined
characteristic is very effective since the instantaneous element produces very fast operation for
high current faults, and the long time delay inverse element provides the desired security and
facilitates selectivity in the low current range.
Relay elements typically have easy to use setting systems. For example, the pickup current of
traditional electromechanical inverse time relays can be changed by changing the position of an
accessible screw. The curve level is adjusted by changing the position of the time dial setting
(TD).
⎡ 5.95 ⎤
t = TD ⋅ ⎢ 2 + 0.18⎥
⎣ M −1 ⎦
I
M=
I pu
This figure shows a family of IEEE Inverse (U2) Characteristics. The numbers on
the curves are the corresponding time dial setting values.
By increasing the TD you also increase the operating time. Depending on the relay
design, this dependency may be directly proportional or not.
In this example, for the element shown, changing the TD from three to nine makes
the operating time three times longer.
1
Radial
2 F1
Radial
3 ~0
z Selectivity Condition:
t2 > t1 for faults at F1
z No Need to Check Coordination Between
3 and 1
For the case shown, there is no need to calculate the setting of Relay 3 to
coordinate with Relay 1 for faults at F1, since the current measured by Relay 3 is
considerably less than Relay 3’s pick-up.
1 ~0
Radial
2
Radial
3 F2
z Selectivity Condition:
t2 > t3 for faults at F2
z No Need to Check Coordination Between
1 and 3
2 1
F1
4 3 F2
z Selectivity Conditions:
t3 > t1 for F1
Contradictory
t1 > t3 for F2
As shown in the figure, there are two selectivity conditions which are often
contradictory. Finding relay settings that can achieve both conditions will be
difficult.
We do not want Relay 1 to detect faults at F2, nor do we want Relay 3 to detect
faults at F1. The condition can be met if the relay is modified to operate only when
a fault occurs on its line. In other words, the overcurrent relay needs to be
directionally sensitive.
2 1
F1
4 3 F2
z Selectivity Conditions:
No Need to Verify if
t3 > t1 for F1
1 and 3 are Directional
t1 > t3 for F2
Overcurrent Relays
If the directional element is designed to operate when the fault is on its protected
line, by detecting the direction of the current, there is no need to verify the
coordination constraints that could cause a contradiction.
Directional overcurrent elements will be the focus of this section; the other
applications will be explained later on the course.
Basic Scheme of a
Directional Overcurrent Relay (67)
V Directional Element
(32)
I
Permissive,
blocking or restrain
I Overcurrent Element
(50 or 51) Trip
The arrows shown in the figure are used to represent the protection tripping direction.
Note that the relays are oriented towards the protected lines. This orientation divides
the system protection into two independent groups: the relays “looking” to the right
and those “looking” to the left. The directionality divides the coordination process
into two independent processes. A relay only needs to be coordinated with the other
relays in its group.
The system shown in the lower figure is a single ring with only one source. In this
case, all the relays are directional except the relays adjacent to the generation bus.
For line faults close to the generation bus, the system is inherently directional. That
is, fault current can only flow out of the bus and into the lines. Thus, there is no need
for a directional relay.
Note: the time dial settings of the relays at locations K and L should be set to the
minimum value. Can you explain why?
I III
II IV
A radial system with two parallel lines is a unique case. In such a case, directional
overcurrent protection is needed at the line ends connected to the load bus. The
addition of a parallel line in a radial system creates the necessity for directional
protection.
1) The relay settings at either terminal are the same. Thus, the settings of I =
settings of II, and settings of III = settings of IV.
2) The time dial setting of relays at III and IV can be set at their minimum.
12 15
17
13
21 20 18
14 16
19
Radial circuit
Legend: Directional OC relay (67)
Non-directional OC relay (51)
One of the first activities to carry out when performing these calculations consists
of establishing the set of coordination pairs. A coordination pair is a group of two
relays, a main relay and a back-up relay. The back-up relay is set to operate in case
the main relay, or its circuit breaker, fails to operate for faults in its zone.
Generally, for directional overcurrent relay coordination, each pair follows the
remote back-up rules described in a previous section. The following table shows
some of the coordination pairs for the system shown in the figure. Review and
complete the table:
Main Back-Up
2 6
2 13
2 17
3 1
3 13
3 17
: :
: :
Directional Elements
Basic Principle
V I
F2 F1
Relay
V I
Now that is has been determined that directional relays are needed, how are they
implemented?
A classical directional element responds to the phase shift between the relay
voltage and current. For faults on the protected line (forward faults), the current
lags the voltage. The angle between voltage and current corresponds to the angle
of the fault-loop impedance.
For faults on the adjacent line, the voltage angle remains almost unchanged and
the current angle changes by approximately 180°. The directional element uses
this information to discriminate between forward and reverse faults.
Observe that the voltage input signal acts as an angular reference. This signal is
referred to as the relay polarizing quantity. The current input signal contains
information about the fault location and is referred to as the relay operating
quantity.
POL
Early directional relays consisted of induction cylinder units. The relay’s operation
is based on the movement of a cylinder. The torque to move the cylinder is the
magnetic flux produced by the interaction of the two input quantities.
This torque equation is very well-known, and modern digital relays still use the
term “torque” to reference the quantity which determines the relay operation.
Solid state analog directional elements were designed and built in the 1960’s and
1970’s. These essentially were based on the angle comparator principle. In
general, the design of these elements followed similar rules to the ones used to
implement directionality in electromechanical relays.
ZTL
k V I cos(θ −θMT ) ≥ 0 θ = ∠V − ∠I
No
Operation
Operation
θ V (Polarizing
θ MT Quantity)
ZTL Maximum I (Operating
Torque Line Quantity)
(MTL)
This type of relay is called a directional relay with a lagging maximum torque
angle.
Va, Vb, Vc
Relay
It has been established that a directional relay needs two input quantities to
determine direction. The angular relationship between currents and voltages is
dependent on the type of fault.
First, let us consider the phase fault relays. The figure illustrates the most widely
used traditional phase directional element. Each of the phase elements receives a
line voltage and the current of the third line as input signals. The phase shift between
relay voltage and current for a unity-power-factor condition is 90°. Thus, the name
of the connection is the 90° or Quadrature connection.
A value of 30° or 45° is selected for the maximum torque angle θMT with current
leading the voltage. For forward fault conditions, the fault current lags the phase
voltage and the unity-power-factor current and tends toward the maximum torque
line, MTA, as shown in the figure. If θMT is 30°, the fault current for a bolted fault
will coincide with the MTA if the characteristic line angle is 60°.
It can be shown that this is valid for three-phase and phase-phase faults. For forward
three-phase faults, all three directional elements will be in close proximity to the
maximum torque condition, and will operate. For forward line faults, there are only
two elements with positive operating torque. The directional element connected to
the voltage between the faulted lines will see no fault current and will likely have a
low polarizing voltage, and will not operate.
For reverse three-phase or line-to-line faults, the relay current will be shifted by
180°, placing it in the restraint region of the directional element.
The operating torque for each of the three electromechanical elements is shown
above. Each relay operates when its corresponding torque is larger than a given
quantity, generally zero.
a
b
c
67+ 67+ 67+
Ia + + +
Ic
Ib
52
Line
The figure shows the typical connection diagram for directional elements using the
90° connection. The manufacturers literature should always be consulted to verify
the proper connections to achieve the desired directionality.
Note that in the connection the polarity markings of the relays, CTs, and VTs are
selected in order to provide the desired tripping direction. For currents flowing out
of the bus and into the line, the relay currents flow into the relay polarity markings.
Accordingly, the voltage drops in the secondary side of the VTs are oriented from
polarity to nonpolarity marking in the relay. The result is positive relay torque for
faults on the protected line for forward faults.
Note that the secondary connections of CTs and VTs are grounded at only one
point. If more than one ground exists, there could be circulating currents during
ground faults that would create a potential difference between the two grounding
points. For example, if there is a ground at the relay panel and the substation
switchyard, the circulating current can cause a relay misoperation.
Polarizing Operating
Quantity Quantity
3V0 3I0
3I0 3I0
Non- V2 I2
Traditional
Zero-sequence directional elements have been used to properly detect ground fault
in power systems. These elements are very sensitive and generally secure.
Whatever the directional element type and connection might be, the overcurrent
element always responds to the residual current 3I0.
V0
I0
+
3I0
I0 V0
+
ZT0 ZL0
3V0
Observe that current 3I0 leads voltage 3V0 by more than 90°. In traditional
directional elements, it is difficult to obtain a value of θMT greater than 90°.
Options
3I0
Voltage Current
3V0 -3I0
-3V0 3V0 -3V0 3I0
-3I0
Example:
θMT = 60° (Lagging)
There are two possible relay connections to keep θMT less than 90°. One solution
uses 3V0 and -3I0 as input signals, and the other connection uses -3V0 and 3I0.
In both alternatives, the operating current lags the polarizing voltage by the angle
of the zero-sequence network impedance. Because of the lagging operating
current, a value of 60° is commonly used for the maximum torque angle θMT, with
the current lagging the voltage.
a
b
c
3I0
3V0
+
+ 67N
52
Line
The figure shows a connection for the 3V0 polarized ground directional element.
According to the polarity markings, the relay receives -3V0 and 3I0 as input signals
in this connection. In actual practice, the manufacturer will likely require the
polarity of the CTs and VTs to be connected to the polarity of the relay. The
manufacturer will then reverse the polarity of one of the inputs to obtain the
desired functionality. As always, you should consult the manufacturer’s literature
before making any connections.
Observe that the residual connection of the CTs is used as a zero-sequence current
filter. In order to get 3V0, a set of auxiliary VTs are connected grounded wye-
broken delta. This is a traditional zero-sequence voltage filter.
Note that the use of auxiliary VTs is not a requirement. The primary VTs can be
connected grounded wye-broken delta to provide 3V0 to the relay, if secondary
phase voltages are not needed. Additionally, many VTs are supplied with dual
secondaries. In such a case, one set of secondaries can be connected as wye to
supply phase voltages and the other set of secondaries can be connected broken
delta to supply 3V0.
These equations summarize the torques for the four electromechanical relays. The
phase relays are in the so-called 90 degrees connection. The formulas can be used
to check the operation of these directional elements for a given fault with known
phase voltage and current values.
If a given relay has a positive torque, the directional relay indicates forward
direction.
Flux-
Summing
CT
I0
b
c
I0
3I0 3I0
3I0
3I0
Previously, it was indicated that zero-sequence current could be used for the
polarizing and the operating quantities of a ground fault directional element. The
question is, “How to use the ‘same’ quantity for both inputs of a directional
element?”
The figure shows three sources for obtaining 3I0. Starting from left to right, 3I0 can
be obtained from a CT in the grounded neutral of a wye-connected winding of a
power transformer. Next is the residual current of wye-connected CTs. Last, 3I0
can be obtained through the use of a flux-summing CT.
S
138kV 67N
S 69kV
S
67N
• All paths for ground current to flow are not necessarily ground sources.
• If they merely pass ground current from another source, they cannot be used for
polarization.
• The H0X0 bushing current of auto banks can only be used with caution. The tertiary
delta current is a better source.
3I0L 3I0T
+ 67N
+
52 3I0L
Line
The figure shows one alternative for 3I0 directional element connection. The
current flow directions are shown for a ground fault on the protected line. The
primary zero-sequence current flows up the transformer neutral and out the line.
The secondary zero-sequence current flows into the relay polarity markings.
For a ground fault on another line connected to the same bus, the primary currents
will flow up the transformer neutral, and into the bus from the protected line. For
such a case, the phase angle for polarizing current 3I0T remains unchanged and the
operating current 3I0L undergoes a 180° phase shift.
138kV 69kV
1I0TTR
IPol=1I0TTR
IPol=(3I0SR+3I0TTR)@138kV-3I0SR@69KV
3I0SR+3I0TTR 3I0SR
An easier alternative is to use the current circulating in the delta for polarization. It
will be a reliable source of polarization in all cases.
If V= 3 V
I = 0.6 A
In this case the relay will not operate, regardless of the angle between the current
and the polarizing voltage.
Directional Element
Issues and Solutions
Relay 1
However, for close-in faults, the voltage may be so low that the minimum operating
torque cannot be overcome. To offset the low voltage, there is an increase in
current magnitude.
The most obvious problem is for a fault at the breaker, where the voltage is virtually
zero.
I Fault
Open
Relay 1 Relay 2
Bus S Bus R
A-Phase
Fault
Relay 1 should see the fault in the forward direction, while Relay 2 should see the
fault in the reverse direction.
44.3° 44.3°
180°
136° 136°
IA , IB , IC
VPOLA=VBC VPOLA
Relay 1 A-phase element sees the fault in the forward direction, as expected. Relay 2
B-phase and C-phase elements see the fault in the forward direction, instead of the
expected reverse direction.
Bus S Bus R
Line 2
Source S Source R
ES ∠ 30° ER ∠ 0°
Line 1
Relay 1 Relay 2
Z S1 = Z R1 = j0.8 Ω Z L1 = j4 Ω
Relay 1 Relay 2
I1 ⋅ 1∠ ZL1
V1
θ = 74°
I1 V1
I1 θ = 86 °
I1 ⋅1∠ ZL1
For Relay 2, the angle between the polarizing quantity and the operating quantity is
less than 90°. Relay 2 incorrectly declares the fault as forward.
The amount of torque is directly related to the magnitude of the voltage. Hence,
zero volts results in zero torque.
The close-in zero voltage fault issue can be overcome by using positive-sequence
memory voltage.
Zero-Sequence Polarization
Parallel Lines w/ Common Terminals
Line 2
ZOM
Line 1
Source 1 Source 2
SLG Fault
Zero-sequence polarized directional elements are reliable for use in parallel lines with
common terminals.
Zero-Sequence Polarization
Parallel Lines w/ Isolated Zero-Seq. Sources
Relay 3 Relay 4
Line 2
Source 3 Source 4
ZOM
Relay 1 Relay 2
Line 1
Source 1 Source 2
SLG Fault
Zero-sequence polarized directional elements can misoperate in parallel lines with high zero-
sequence mutual coupling and isolated zero-sequence sources.
The figure shows parallel lines with isolated zero-sequence sources and a SLG fault on Line
1. Note that the zero-sequence current in each source is in phase with the zero-sequence
current of the corresponding relay. An undesired trip may occur in the unfaulted line.
Zero-Sequence Polarization
One Common Bus
Relay 3 Relay 4
Line 2
Source 3
ZOM
Relay 1 Relay 2 Source 2
Line 1
Source 1
SLG Fault
Relay 3 Relay 4
Line 2
Source 3
ZOM
Relay 1 Relay 2 Source 2
Line 1
Source 1 Open
SLG Fault
Zero-sequence source isolation also occurs in parallel lines with a single common bus after
the breaker closest to the common bus opens.
Summary
z Directional elements ensure the speed,
security and selectivity of non-unit
protection schemes in looped systems.
z The most common scheme for traditional
directional elements includes three phase
fault relays in 90° and one ground relay.
z Directional overcurrent relays’ settings
must be calculated to achieve
coordination.
Summary