LG_PRO 001
LG_PRO 001
LG_PRO 001
LEARNER MODULE
1 INTRODUCTION ................................................................................................................................................... 2
1 INTRODUCTION
This training module looks a little at the clearance of transient electrical faults on overhead
lines, though much greater emphasis is placed on the subsequent automatic restoration of
power flow after such faults have been cleared. Overhead lines are used mainly in Medium
Voltage (MV) Distribution Systems (1kVac – 20kVac), High Voltage (HV) Subtransmission
Systems (20kVac – 200kVac) and also Extra High Voltage (EHV) Transmission Systems
(200kVac – 1000kVac). (UHV>1MV).
The main reason for this is that overhead lines are much cheaper to construct and maintain
than cables. The biggest cost saving is achieved by the reduced amount of insulation
material used on an overhead line. On an overhead line the conductors are insulated from
ground potential at the tower structures by means of string insulators that are constructed of
glass discs or alternatively porcelain or silicon rubber insulators. Apart from that the
conductors are bare, and are otherwise insulated by the ambient air.
Air is capable of restoring itself naturally, provided that the source providing the fault current
is isolated. This gives another advantage that overhead lines enjoy compared to cables,
since cable insulation cannot restore naturally once punctured by an electrical fault. Due to
this advantage, the majority of faults (85-95%) on overhead lines are of a transient nature.
By definition, transient faults are cleared when the relevant circuit breakers are tripped and
do not recur when the line is re-energised.
Faults on cables are very rarely transient, since the damage caused by the fault puncturing
the cable insulation material raises a very high probability that it will fault again when re-
energised. For all practical purposes, a fault on a power cable is permanent. Thus,
overhead lines provide the advantage of transient faults. However, for this advantage to be
of benefit, Automatic Re-Closing (ARC) is employed, to re-energise the line as soon as the
fault has been cleared, so that power flow can be restored.
In a distribution network it means that the customer’s electrical supply is interrupted
temporarily for a short period, and then restored by means of ARC. In a transmission
network it means that the network power flow will be rerouted through other transmission
lines for a short period while the local (faulted) line is de-energised, and then the power flow
will be routed through the local line again as it is re-energised by means of ARC. This might
lead to temporary overload conditions on some transmission corridors, but since it is for a
short period, it rarely poses a risk to plant damage. It is also for this reason that overload and
overcurrent protection is normally not used on transmission lines.
ARC provides benefits for both the customer and the supplier. When a faulted line trips,
there is an increased source impedance to the customer which can result in a prolonged
voltage dip at the terminals of the customer’s equipment. ARC therefore improves the
customer’s quality of supply by reducing the duration and magnitude of voltage dips. ARC
also improves the security of supply for customers with multiple infeeds, and increases the
availability of supply for those with a single infeed by minimising the duration of the
disruptions. As far as the supplier is concerned, ARC enhances stability on the
interconnected system by reducing the connecting impedances between sources.
There are however several adverse conditions that prohibit ARC, or at least limit ARC to only
particular operating conditions. There are a few different selections that can be made for
ARC schemes, and this allows different basic philosophies for distribution systems and
transmission systems. Differences in philosophies are due to the different switching
equipment that is used and their respective capabilities, as well as different customer
requirements or supplier requirements. This training module will discuss the above ARC
Compiled by Hans Bekker 2
Manual for Short course on overhead feeder Auto Re-close principles and practices:
Rev 2
philosophies and explain particular settings for ARC schemes as well as special cases that
require deviations from the basic philosophies.
2 CONDITIONS THAT PROHIBIT ARC
When a fault is of a permanent nature no ARC should take place. Reclosing back onto a
fault could be catastrophic in certain circumstances. In most cases it is rather difficult (or
maybe impossible) to distinguish from measurements whether a fault is transient or
permanent, and since most faults are transient, ARC can be applied at all times based on the
stochastic probability that the fault would have been transient. However, cable faults are
almost without exception always permanent, so there is no ARC on a faulty cable.
Likewise, no ARC is to take place for busbar faults or transformer faults. Reclosure onto a
fault in a substation could cause equipment to explode or start a fire. Prevention of further
damage to the equipment in the substation is to be considered simultaneously with safety of
personnel that may be working in the substation yard.
Sometimes a line fault and a busbar fault can occur almost simultaneously. ARC should be
inhibited whenever a busbar fault is cleared, even if the feeder protection tripped for the line
fault first. That is to say that whenever a busbar trip is issued, that trip signal should also be
used to block any ARC relays on the circuits that it trips, and it should even block ARC relays
that have already been started.
No ARC is to take place for transformer faults, since a reclosure could ignite leaking
transformer oil. Similarly, ARC should not take place after faults on the transformers of Line-
bank-transformer feeders. The Line-bank-transformer differential protection should operate
for a transformer fault and transmit a signal for an ARC lockout at the feeder. For a carrier
transmit or a receive failure or a differential protection failure a reclosure could ignite leaking
transformer oil.
In addition, on Line-bank-transformer feeders it is not possible to clear a fault by means of 1-
phase tripping, since the magnetic coupling through the transformer core will sustain the
fault. This necessitates 3-ph ARC for all fault types on Line-bank-transformers, and thus the
transformer is de-energised and re-energised in short succession with every fault. The
increased cycle of the accompanying inrush currents may shorten the serviceable lifespan of
the transformer.
In general it is preferred that there should be no ARC at all on line bank transformer
schemes. Where system stability absolutely requires ARC, a Line-bank-transformer may be
selected the 3-ph ARC, provided that the transformer differential protection will inhibit ARC
when it operates.
During operating, a control centre might determine that there are adverse conditions (such as
widespread veldfires) that make many repeated (unsuccessful) ARC attempts (with their
accompanying voltage dips) undesirable. During such conditions the control centre might
override any ARC selections and switch ARC off on affected lines.
In the Chapter titled: Examples of Exceptions, there will be several more specific cases that
require ARC to be off or limited to certain modes. Each case has several conditions, so that
a general rule cannot be formulated, but each has to be evaluated on its own merit.
A cycle will continue in sequence if a recurring fault occurs within the reclaim time. The
reclaim time is usually factory set to 15 seconds, however, if the ARC relay allows, Eskom
has decided on setting a reclaim time of 240 seconds or four minutes, to allow the circuit
-δ +δ
5.4.3 Standard Dead Line Charge End Selection for Transmission Systems ARC
The standard Transmission feeder ARC Dead Line Charge End selection is determined by
the “Protection Setting Philosophy for Transmission and Sub-Transmission Grids”, and also
the “ARC Philosophy for Lines with Shunt Reactors”. A summary is given below, subject to
reference with the abovementioned Philosophies:
The selection of dead line charging end of the line is driven by three requirements:
The dead line charging end should provide sufficient fault current for the relays to operate
reliably.
The dead line charging should not result in excessive voltage depression.
The dead line charging should not be selected at the power stations.
The dead line charging end is normally selected in line with the Table below based on the
fault level (FL) at the connected substations A and B. The fault level has to be calculated with
the line A-B taken out of service to simulate the fault current that will be measured by the
relay.
Substation B:
Substation FL<10kA Substation with Substation A Substation B
higher Fault Level
Substation FL>10kA Substation B Substation with Substation B
lower Fault Level
Power Station Substation A Substation A Power station with
lower Fault Level
Before selecting the DLC end the design of ARC circuits on both ends of the line must be
checked. Some very old protection panels are not equipped with synchronising facilities.
Such line end must be selected as the DLC end.
For a substation having only two lines fed from different sources, no synchronism check is
possible if a fault occurs on the one line during an outage on the other. One available option
is:
Select one end to be a dedicated synchronising end. If the other line is out at the time,
the breaker at the synchronising end will have to be closed in manually (possibly via
supervisory)
On a radial feeder (a section of network which has only one station as the source of supply),
perform a dead line charge from the source end, even if it is a power station. At the remote
end, a live line/dead busbar close must be allowed, but still with a longer delay time than the
charging end. For a parallel line radial feeder, force a synchronism check at the remote end
of the lines. This ensures that, for a sustained fault on one line with the parallel line in
service, reclosure at the remote end onto the sustained fault will not be possible. However, if
the parallel line is out of service, manual closing of the remote breaker is necessary to
restore the line to service.
For cases of only marginally different fault levels, charge North – South.
4. OFF selected. The table below indicates the ARC sequence for a sustained line fault.
FAULT TYPE 3 TRIP ARC L/O
1 FAULT X X
MULTI FAULT X X
6. 1 POLE ONLY selected – not possible on 3 pole only schemes. The table below
indicates the ARC sequence for a sustained line fault.
FAULT TYPE 1 TRIP 1 ARC 3 TRIP ARC L/O
1 FAULT X X X X
MULTI X X
FAULT
A cycle will continue in sequence if a recurring fault occurs within the reclaim time. The
reclaim time is usually factory set to 15 seconds, however, if the ARC relay allows, Eskom
has decided on setting a reclaim time of 240 seconds or four minutes, to allow the circuit
breaker components to restore and cool down. Once the reclaim time has elapsed, any fault
then occurring will cause the sequence to begin again.
6 EXAMPLES OF EXCEPTIONS
6.1 ADDITIONAL ARC REPETITIONS
On several power flow corridors a transmission grid’s network configuration may require that
additional ARC attempts should be made, to increase the probability of successful load
resumption. Some are due to voltage stability conditions, and others are due to unfirm
supply conditions. In these cases the ARC selection is made to 5-cycle ARC, which will
attempt one extra 3-ph-ARC after an unsuccessful 1-ph-ARC attempt.
8.3 WHY USE NEUTRAL REACTORS ON EHV LINES WITH SHUNT REACTORS?
First, let us consider a transmission line without a shunt reactor, and let it trip single phase for
an earth fault from one phase to ground. As can be seen, there is no significant path for the
fault current to flow along once the circuit breakers at the two ends have opened. The only
problem that might occur is unacceptable high voltages at an open end if the other end
closes before it. This would especially be a problem if one end tripped three phase while the
other end tripped single phase.
Figure 8.3.1: Single phase tripping transmission line without shunt reactor
However, let us assume this is a 300km long line, and to solve the Ferranti Effect, we have to
add a shunt reactor. Let us consider what happens during a single phase trip.
Figure 8.3.2: Single phase tripping transmission line with shunt reactor
Now it can be seen that even while the circuit breakers for the faulted phase are open, there
is still a clear route for fault current to flow along. All that is needed is some capacitive
(and/or inductive) coupling from the other two live phase conductors, and there will be
sufficient EMF to support the fault current indefinitely. On a 300km transmission line, the
mutual capacitive and inductive coupling between the phases are enormous, and we can
expect that once the faulted phase circuit breakers ARC, the fault will still be there, and the
protection will have to issue a second trip. After this trip the protection will not allow a second
ARC attempt, and our hopes to restore the power system quickly and automatically will be
dashed to pieces.
So in order to stop the current flow along this earth path, we insert a high impedance neutral
reactor. The impedance value of this neutral reactor is carefully selected so that it will be
high enough to reduce the fault current significantly, and thereby cause it to break
spontaneously during a zero-crossing (thanks to AC).
Xnr
Compiled by Hans Bekker 19
Manual for Short course on overhead feeder Auto Re-close principles and practices:
Rev 2
Figure 8.3.3: Single phase tripping transmission line with shunt reactor, and tuned neutral
reactor that causes fault current to be disrupted during a zero-crossing
Unfortunately, this additional impedance between the reactor neutral and earth also raises
the standing voltage that is superimposed on the isolated (supposedly “Dead”) conductor
through capacitive coupling.
In fact, without a line reactor, there is also a voltage superimposed on the isolated conductor,
although the forces that support that voltage are then quite weak. With the line reactor
directly earthed, this voltage is much reduced, but when a neutral reactor is put between the
main reactor neutral and earth, the voltage on the isolated conductor is increased by a large
factor again. Particularly for this case the neutral reactor actually provides a strong
connection through which that voltage can be supported.
Due to this voltage on the “Dead” conductor, there is always a compromise that must be
reached when the neutral reactor impedance is designed. On the one end, there is the
probability that the secondary arc will never clear. As the neutral reactor impedance is made
larger and larger, this probability reduces. On the other end, there is the probability that a
new secondary arc will be ignited by the recovery voltage over the weakened air insulation.
This probability increases with the impedance of the specified neutral reactor. Figure 8.3.4
illustrates this compromise of probabilities graphically, where the thick red graph indicates the
probability of Auto-Re-Close failure. The optimal selection for the neutral reactor impedance
is where this graph reaches a minimum.
Z(neutral reactor)