Power System Protection: Bus Protection
Power System Protection: Bus Protection
Power System Protection: Bus Protection
CONTENTS
OBJECTIVES
The objective of this note is to review the different bus layouts that are used in power systems, and
to present the different protection schemes that are installed to protect against bus faults.
After study of this note the participant should be able to understand the following overall concepts
and apply them to his day-to-day work activities. He will also be able to answer test questions on
these subjects:
SEGMENT A: - BUS ARRANGEMENTS
The bus is a very important part of the power system, located at every junction point of lines and
feeders. Invariably, a high voltage bus is constructed in the open air, with the conductors consisting
of solid bars or tubes. Air provides the insulation between the conductors. At various points along the
bus, transmission lines or other circuits are connected through circuit breakers.
At lower voltages the bus may consist of solid copper bars running along the back of metal clad
enclosed switch gear which can be built indoors or outdoors. Common types of faults that can occur
on a bus are:
Faults rarely occur on buses, but if a fault does occur, then it must be cleared as quickly as
possible, because of the potential for extensive damage to equipment.
As shown in Fig. A-1, the direction of current flow forms the basis of bus protection. To measure
the total flow of current into and out of the bus, CTs are located on the outboard side of the circuit
breakers. If there is no fault on the bus, then the sum of the currents leaving is equal to the sum of
the currents entering; the total is zero. Even if there is a fault on one of the transmission lines
downstream, the sum of the currents is still zero as the total current entering the bus is matched by
that leaving. The downstream fault would be cleared by other protection relays activated by CTs on
the inboard side of the circuit breaker.
If a fault occurs on the bus itself, the sum of the currents in and out of the bus will no longer be zero.
This will cause the protection relays to operate and trip all the breakers connected to the bus.
Single Bus
The most simple bus arrangement is the single bus shown in Fig. A-1. The protected area is defined
by the location of the CTs. The inboard CT on each breaker is used for protection of the connected
circuit. The CTs are usually located in the circuit breaker bushings. To simplify the diagrams, the
breaker disconnects are shown in the closed position, unless they are normally open.
Split-Bus with Tie Breaker:
In order to improve reliability, the bus is sometimes split with a tie breaker, connecting the two
halves as shown in Fig. A-2. In this arrangement, the tie breaker will normally be closed in order to
increase system reliability. Each side of the bus, including the tie breaker, is protected by separate
relays, so providing two separate zones of protection.
A fault occurring on bus B would cause the connected circuit breakers, including the tie breaker, to
open. This would isolate the faulty half of the bus, but leave the remaining section, bus A to operate
intact. As there is only one circuit breaker in each circuit, when a breaker needs to be isolated for
maintenance the line has to be taken out of service. One economical way of getting around this
problem is to provide a transfer bus, as shown in Fig. A-3.
Double Bus:
The double bus arrangement is shown in Fig. A-4. The essence of this arrangement is that
each circuit can be connected to either bus A or bus B according to the selection of the circuit
disconnects.
Differential protection is provided separately for each bus, as shown in the diagram.
This configuration allows us to remove a breaker from service without interruption. Fig. A-5
shows the arrangement with breaker 1 out of service, and bus A functioning as the transfer bus. Note
before closing the transfer breaker, differential protection on both buses must be disabled since
closing the transfer breaker unbalances differential protection on the buses.
With this arrangement it is necessary to provide two CTs for line protection, one on each alternate
breaker connected in parallel. The VTs are connected one to each line.
Ring Bus:
Another bus arrangement commonly used is the ring bus, shown in Fig. A8. Circuits are
connected to a common ring, which in turn is broken into sections by circuit breakers. A fault on any
line will cause the tripping of the two breakers on either side.
With this arrangement, special bus protection is not required as each bus section is included
in its associated line protection. If a circuit breaker is removed for maintenance, the bus still remains
connected to the four circuits but it is now an open bus. However, in this situation, if a fault occurs on
one line, then the bus will be split as additional breakers open.
Directly Connected:
Fig. A-9 shows a circuit connected to a bus directly through a transformer (there is no breaker
between the transformer and the bus). In this case, the bus zone protection is extended to include
the breaker and outboard CT on the primary side of the transformer. A fault on the bus or in the
transformer will operate the bus protection to trip all the connected breakers including the
transformer primary.
Alternately separate protection could be installed on the bus and the transformer.
Transformer faults would trip all the breakers as before. Then a motor operated disconnect switch
installed between the transformer and the bus would be opened and the bus put back into service.
The simplest method of providing differential protection is to fit CTs to measure the current in all
circuits connected to the bus. For clarity, Fig. B-3 shows one phase only of each line. The secondary
side of the CTs are all connected in parallel and the residual current (that is, the current difference)
passes through an overcurrent relay. If the current in and out is balanced, then the residual sum of
the secondary currents in the CTs is zero.If there is a fault on the bus, then some or all of the circuits
will feed into the bus. The current in the CT secondaries will now be additive and the resulting
residual current will operate the time overcurrent relay.
This means that for the first few cycles after initiation of the fault, the output of the heavily
loaded CT will be most unreliable and will probably cause false tripping.
The most simple way to get around this problem is to use the delay of the time overcurrent
relay to ride through the transient period. A time delay of about 0.2 seconds (that is 10 cycles) is
sufficient on most applications.
More sophisticated arrangements are needed for protection of high voltage buses. One
improved method uses directional relays fitted to every circuit connected to the bus (Fig. B-6). The
operating contacts from the relays are all connected in series. When current flow is out of the bus, its
respective relay contact is held open. Conversely, when current flow is into the bus, the contact is
closed. For a bus fault, current will flow into the bus from all circuits, and, consequently, all of the
contacts will be closed, energizing the relay to trip all of the breakers. Where radial feeders are
connected, there would be no reverse current into the bus and, therefore, these are not included in
the scheme.
Fig. C-1 shows a numerical example of a simple current differential. For simplicity we show
just one phase on each line. Assume the CT ratios to be identical at 100 to 1 for a rating of 500 amps
primary. Fig. C-1 shows the magnitude of current and its direction of flow in each of the CT
secondaries. The secondary currents cancel out and there is no current passing through the relay
operating coil as the loads are balanced.
Fig. C-2 shows the situation when a fault occurs on the bus with a magnitude of 18,000
amps. Now we will have all six lines feeding current into the bus to supply the fault. Depending upon
the reactance of each line from its respective source, the current flows will probably be similar, in this
case about 3,000 amps each. In the secondary of each CT we have 30 amps flowing. This is six
times the secondary rating of 5 amps, well within the range of accuracy (you will remember that
protection type CTs retain 90 per cent accuracy up to twenty times rated current). These secondary
currents are all additive so we have 180 amps passing through the relay operating coil. This will
operate and trip all of the bus breakers. This is correct for an internal fault.
Fig. C-3 shows a fault on line 4 fairly close to the breaker. As before, the fault magnitude is
18,000 amps. Lines 1, 2, 3, 5 and 6 will each feed 3,000 amps into the bus and then 15,000 amps
out along line 4 to supply the fault. Line 4 also feeds 3,000 amps to the fault. (Total fault current is
15,000 + 3,000 = 18,000 amps)
The diagram shows the currents in the CTs secondaries. We should have 30 amps on each
of these lines and minus 150 amps on the faulty circuit. This would give us zero current passing
through the relay, hence it will not operate. This is correct for an external fault.
But in practice the relay will probably operate incorrectly as the level of current in the faulty
line is 30 times its rating, hence the CT is over its saturation limit. It is likely that the secondary
current will be, say, 110 amps instead of 150. So now we have a current of 40 amps passing through
the operating coil of the relay, sufficient to cause a trip. This is incorrect operation; we do not want
the bus protection to operate for external faults.
One method to improve the reliability of operation during this transient period is to utilize a
differential relay with multiple restraint coils, as shown in Fig. C-4. We have a total of 200 amps
restraint in one direction and minus 60 amps in the other. Therefore total restraint is proportional to
140 amps. In this case the percentage differential is the operating coil current 40 amps divided by
the total restraint current 140 amps, which equals approximately 29%. In order to allow high speed
tripping and avoid misoperation during the first few cycles of the fault, the percentage differential
would be set to about 50%.
The multi-restraint relay is quite fast acting; it operates within 50 to 100 milliseconds, that is,
approximately 3 to 5 cycles.
In order to improve accuracy, it is important that the burden on CTs be kept as low as
possible; therefore, the CTs should not be used for any other purpose. For the same reason the
leads connecting the CTs should be of ample current carrying capacity. As all of the CT ratios are
normally identical, the bus differential relays are not fitted with ratio taps. If a particular circuit does
have a different CT ratio, then it is necessary to connect an auxiliary CT in order to match the others.
Another way of getting around the problem of CT saturation with bus protection is to use a
different type of CT, one which has no iron core. The air core CT is known as a Linear-Coupler. The
Linear Coupler is usually designed to provide 5 volts on the secondary side for 1,000 amps passing
through the primary. The type of relay used is a voltage differential relay.
For bus protection, the linear couplers are all connected in series, as shown in Fig. C-5.
For a fault on the bus, lets say 8,000 amps, each of the transmission lines will feed in 2,000
amps. We will now have a total of 40 volts across the relay operating coil and the relay will trip. This
is correct.
Figure C4: Simple Current Differential Circuit with Restraint Coils and a Close in Line Fault
Figure C5: Bus Protection with Linear Couplers
For an external fault on one of the lines, we may have, say, 3,000 amps entering in each of
the lines 1, 2, and 3 and 9,000 amps going to the fault on line 4. As the linear coupler does not suffer
from saturation effects the sum of the voltage will be zero. The relay will not trip for this external fault.
This relay is highly sensitive and will trip in about 20 milliseconds, that is, one cycle, thus providing
high speed operation.
SEGMENT D - HIGH IMPEDANCE DIFFERENTIAL RELAYS
High speed differential bus protection is often provided by the use of high impedance relays,
which are designed to eliminate the problem of CT saturation discussed earlier.
Fig. D-1 shows the arrangement. A resonant circuit is connected in series with the relay
operating coil. This is a capacitor and reactor which are tuned to allow the passage of 50 hertz
current only. It effectively filters out any harmonics and DC components due to wave distortion as
occurs during the initial transient. One problem with this arrangement is that, when the setting of the
relay coil is adjusted, it changes the inductance in the circuit and thus changes the resonant
frequency. In order to avoid this, a rectifier is placed in the circuit so that only DC current is fed to the
relay operating coil.
The effect of the rectifier and the resonant circuit raises the total impedance across the relay to
about 2,500 ohms. This impedance is far, far greater than the impedance of the CTs, which is
probably about 2 ohms or less. Any error current which exists as a result of saturation, or different
CT characteristics, will flow through the CTs rather than through the high impedance relay circuit. To
allow the correct distribution of this error current it is important that the resistance in each of the CT
branches be equal and as low as possible up to the junction point.
When there is an internal fault on the bus, all of the circuits will feed into the bus and,
therefore, all of the secondary currents will be additive. In this situation all of this current must flow
through the differential relay and operate the tripping circuit. An out of balance current of, say, 20
amps flowing through the CT circuit which has an impedance of, say, 3 ohms, would produce a
voltage of 60 volts across the impedance relay. To trip for this condition the relay would be set at 60
volts. See the next page for a more detailed look at setting the relays pickup voltage.
However, this does raise another problem. If all of this 20 amp current were to pass through
the high impedance relay of 2,500 ohms, the total voltage drop across the relay would be 2,500
times 20, or 50,000 volts. This would obviously destroy the relay. This problem is resolved be placing
a parallel circuit across the relay, so as to by-pass most of the current. This circuit is actually
composed of a varistor (also known as a thyrite resistor), which has the property of decreasing its
resistance as the voltage across it rises.
Light bus faults are cleared by the high impedance relay within 30 milliseconds, that is, about
1 and a half cycles. This is also extremely fast operation. The setting of the high impedance relay is
achieved by selecting the correct voltage tap, usually between 50 V and 500 volts. Lets look at an
example to see how the relays pickup voltage is determined. For an external fault, the CT in the
faulted line may saturate. A voltage (which is the product of the current from the source CT times the
resistance of the windings and leads of the saturated CT) is developed across the bus differential
relay. Lets look at a situation where the worst case or highest saturation voltage that could be
developed across the relay is 100 volts. We do not want the relay to trip at 100 volts, so we would
desensitize the relay by setting it with a safety factor of two or 200 volts. The safety factor would
have little effect for internal bus faults.
Fig. D-2 shows a supervisory circuit which allows the operator to check the integrity of the
high impedance bus differential protection circuit by applying a low voltage of about 40 V. When on
standby, contact A is closed and the 40 volts test voltage is available (This is not enough to operate
the relay). Contact B is closed to check for short circuits. A short in the CTs or the relay causes the
voltage to fall. To check for open crcuits, leave contact B closed and open contact A. If there is an
open circuit in the CT, a voltage will be present. The indicated voltage should be zero if the circuit is
O.K.