Pages

Wednesday, April 16, 2014

Power System Protection Course- GENERATOR PROTECTION





GENERATOR PROTECTION



CONTENTS
GENERATOR PROTECTION...........................................................................
INSULATION FAILURE......................................................................................
EARTHING BY RESISTOR..................................................................................
EARTHING BY TRANSFORMER.......................................................................
STATOR PROTECTION........................................................................................
EARTH-FAULT PROTECTION............................................................................
ROTOR EARTH-FAULT PROTECTION.............................................................
UNSATISFACTORY OPERATING CONDITIONS.......................................
UNBALANCED LOADING.................................................................................
OVERCURRENT PROTECTION.........................................................................
OVERLOAD...........................................................................................................
FAILURE OF PRIME MOVER.............................................................................
LOSS OF FIELD...................................................................................................
OVERSPEED........................................................................................................
OVERVOLTAGE.................................................................................................
PROTECTION OF GENERATOR/TRANSFORMER UNITS....................



GENERATOR PROTECTION

The a.c. generator needs protection against a number of conditions some of which require immediate disconnection and some that rnay be allowed to continue for some time.  In broad terms the former are connected with insulation failure whilst the latter are generally associated with unsatisfactory operating conditions.

Of all the items of equipment which make up a power system the generator is uniquc in that it is usually installed in an attended station and is therefore subject to more or less constant observation.  The point here is that some of the unsatisfactory operating conditions could be dealt with by an operator whereas if the generator was not attended tripping would be the only course of action.

INSULATION FAILURE

Stator faults are caused by the breakdown of the insulation between the armature conductor and earth; between conductors of different phases or between conductors of the same phase.

The most likely place for an earth fault to occur is in the stator slots.  Arcing will probably occur resulting in the burning of the iron at the point of fault and welding the laminations together.  Replacement of the faulty conductor may not be very difficult but the damage to the core cannot be ignored as the fused laminations could give rise to local heating.  In severe cases it may be necessary to dismantle and rebuild the core which is a lengthy and costly process.

To reduce the possibility of damage earth-fault current is usually limited by earthing the generator neutral point via a resistor, reactor or transformer.  Practice varies as to the value to which the current is limited.  From rated current in some cases to very low values in others.

EARTHING BY RESISTOR

Earthing by means of reactors is uncommon and earthing by transformer is usually limited to large machines.  In an industrial system the generator, which is usually directly connected to the power system without a transformer, is earthed by a resistor which has a fairly low value.  The earth-fault current is usually limited to between 50% and 200% of the rated current.

In cases where the generator is connected to the distribution system via a generator transformer a resistor designed to allow an earth-fault current of about 300 A is used irrespective of generator rating.

EARTHING BY TRANSFORMER

The other approach to earthing is to limit the current to a level where burning does not readily occur.  This level is said to be 5A.  To achieve this high-impedance transformers have been used.  Initially voltage transformers were used operating at a fairly low flux density but overvoltage problems arising from the capacitance of the stator windings has resulted in the general use of distribution transformers.  The secondary winding is loaded with a resistor so that under earth-fault conditions a maximum of 5A will flow.
Phase-to-phase faults are far less likely than earth faults and, as they are easily detected, damage caused can be limited by rapid disconnection.  On the other hand, interturn faults, which are also uncommon, are very difficult to detect and are generally only detected and cleared when they have developed into an earth fault.

STATOR PROTECTION

Differential protection using high-impedance relays is usual for stator protection and is applied on a phase-by-phase basis.  As the leads between the two sets of current transformers may be long the resistance will be fairly high but as the maximum through-fault current will be less than 10 times full-load current a reasonably low voltage setting can be applied.  This means that the CT magnetising current will be low and therefore a low overall current setting can be expected.
The overall setting has a direct bearing on the amount of the generator winding which is protected.  This can be calculated as follows:
Max.  fault current-say 5 x CT rating.
Overall protection setting  say 6%.
Amount of winding protected


6%
5
 
 
100% -               = 98.8%.

This would be for a phase-phase fault.  For an earth fault where the current is limited to the full-load value only 94% of the winding would be protected.  In fact slightly less as the full-load current of the generator is usually less than the CT rating.

If the required voltage setting was high because of, say, long CT leads or if the CT magnetising current was high then the overall current setting may be much higher than 5%.  This means that the amount of generator winding protected is also reduced maybe to an unacceptable level for earth faults.  In this case a biased differential relay would alleviate the position.

The use of a biased relay means that the relay-coil circuit impedance can be reduced to about a twentieth of the impedance of the relay coil in the unbiased scheme.  This naturally reduces the voltage setting and the CT magnetising current at setting resulting in an overall setting of about 5%.



FIGURE 7.1  BIASED DIFFERENTIAL PROTECTION.  ONE PHASE ONLY SHOWN

The biased differential scheme is shown in Figure 7.1 and the value of the stabilising resistor, Rs, can be calculated from


RCT + RL + ½RB
B
 
 
RS =                                                    

where B is the ratio of bias coil turns to operate coil turns and is known as the bias ratio and RB is the resistance of the bias coil.

EARTH-FAULT PROTECTION

Where the maximum earth-fault current is restricted to a fraction of the generator rating earth-fault protection is essential to compliment the differential protection scheme.

This earth-fault protection frequently comprises an instantaneous relay having a setting of 10% to 20% and the IDMT relay with a setting of 5% to 10%.  Both relays would be connected to a simple current transformer having a primary current rating equal to that of the earthing resistor.  Earth faults will be detected in 90% to 95% of the generator winding even though the maximum earth-fault current may be as low as 5% of the generator rating.

Even where the main differential protection scheme is expected to provide adequate protection for earth faults an IDMT relay, con­nected to a current transformer in the generator neutral-earth connection, is used to provide back-up protection.  Where the generator is directly connected to the power system, i.e.  without a generator transformer, it provides back-up protection for the busbars and the whole system.  In this case it should have a very long time delay and should be thought of as the last line of defence.

ROTOR EARTH-FAULT PROTECTION

The field system of a generator is not normally connected to earth and so an earth-fault does not cause any current to flow to earth and does not, therefore, constitute a dangerous condition.  If a second earth-fault occurs a portion of the field winding may be short-circuited resulting in an unbalanced magnetic pull on the rotor.  This force can cause excessive pressure on the bearings and consequent failure or even displacement of the rotor sufficient to cause fouling of the stator.  The overheating in the rotor can cause deformation of the winding which could lead to the development of short-circuits.

Two main methods are used for detecting earth-faults in the rotor circuit.  In the first method a high-resistance potentiometer is con­nected across the rotor circuit the centre point of which is connected to earth through a sensitive relay (see Figure 7.2).  The relay will respond to earth faults occurring over most of the rotor circuit.


Figure 7.2  -  ROTOR EARTH FAULT DETECTION - POTENTIOMETER METHOD



Figure 7.3  -  ROTOR EARTH FAULT DETECTION - NEGATIVE BIASING METHOD

There is, however, a blind spot at the centre point of the field winding which is at the same potential as the mid-point of the potentiometer.  This blind spot can be examined by arranging a tapping switch which when operated shifts the earth point from the phase rotation, produce a magnetic field which induces currents in the rotor at twice the system frequency.  This causes considerable heating in the rotor and would cause damage if allowed to persist.

UNSATISFACTORY OPERATING CONDITIONS

These conditions in general do not require immediate disconnection and, it could be argued that, in an attended station the operator could take the necessary action to remove the condition.  Undoubtedly this is possible in some cases but on no account should protection be omitted on this basis.

UNBALANCED LOADING

Unbalanced loading of the generator phases results in the production of negative phase sequence (NPS) currents.  These currents, which have a phase rotation in the opposite direction to the normal

Each generator will have a negative phase-sequence rating which can exist continuously without damage, typically 0.15 p.u. of generator FL current, and an I2t rating when the current exceeds the continuous value, typically I2t = 20.
Where I is per unit NPS current and tis the time in seconds e.g.  the generator would carry a NPS of current 15% full-load current continuously and NPS current of 30% full-load current for a time of


20
0.32
 
 
0.32t = 20                    t =                = 222s.

In fact the time would be longer than the calculated value as there would be some heat dissipation.  An I2t value assumes no heat dissipation and therefore the longer the time the more inaccurate the result.  The result will be fairly accurate up to 2 minutes.

The actual negative phase-sequence current is difficult to deter­mine from the ammeters measuring the load current in each phase.  It is not greater than 65% of the unbalanced current.

Relays to detect the condition usually have an IDMT characteristic matched to the I2t value.  The relay is connected to a network which separates the positive and negative phase-sequence currents.  The basis of the network is to produce a phase shift of 60° in some components of the phase currents such that when the phase rotation is positIve, i.e.  r, y, b, r, the net current in the relay is zero.  When the phase rotation is negative, i.e.  r, b, y, r, a proportion of the current flows in the relay.  Any current which flows in the generator neutral is known as zero-sequence current and this must be eliminated if the network is to function correctly.  Where the generator is connected to the system via a delta/star transformer any zero-sequence current means that there is a fault on the generator circuit and this will be cleared by earth-fault protection.  If the generator is directly con­nected then zero sequence is eliminated by connecting in delta the current transformers which feed the NPS network.  In this case the relay setting is related to the CT current x 1.73.
There is sometimes a reluctance to apply NPS protection as all generators will be subject to the same conditions and could lead to all generators tripping at the same time.  An early warning of the condition can be provided by an instantaneous relay connected to the NPS network to provide an alarm after a short fixed time delay.

OVERCURRENT PROTECTION

An IDMT relay is generally used as back-up protection but the operation of this relay is complicated because of the current decre­ment in the generator during fault conditions.  In some cases a setting is chosen, such that the relay will not eperate for a system fault but will only respond when fault current is fed into the generator, in this way it only acts as a back-up to the main generator protection.

In most industrial installations the relay is required to act as back-up to the system protection and settings must be chosen to ensure positive operation.

The operation of IDMT relays under generator decrement conditions can be calculated by dividing the decrement curve into a number of zones of width, say 0.1 s.  The mid-ordinate is the current level which is converted to a multiple of the relay setting and the time for full travel determined

The difficulty in application arises from the variation in the current decrement depending on generator conditions prior to the fault.  From a no-load condition the current will decay to less than full-load current whereas from the full-load condition the final current will be greater than full-load current because the field current is higher.  The former case will be modified if there is a voltage regulator as this will attempt to boost the field with a consequent increase in final current.  This would have a large effect on the relay and therefore a normal IDMT relay is generally unsatisfactory.  However, this method can be used to determine settings of feeder and transformer IDMT relays in finite busbar systems.  For example, in off-shore installations or any location where the only supply is local generation.  The multiples of setting current in this case will be much greater because the feeder and transformer rated current will only be a fraction of that of the generator.  The higher multiples of setting means that the effect of the difference in generator decrement between no load and full load will be small.

It may be that the current will decay to a level where it is insufficient to cause the overcurrent relay to trip.  In these circumstances it is necessary to provide a relay which not only responds to current but also to the level of voltage.

The principle of operation is that an IDMT relay with a setting much less than the full-load current of the generator has a feature added which increases the setting to above full-load current when full system voltage is present.

By this means the longer operating times, for discrimination with system protection when the fault is remote, will be attainable as the voltage is high.  Close-up faults will remove the voltage restraint to enable the relay to operate in the relatively fast time appropriate to the lower setting.
The relays for this type of protection can be either voltage restrained, where the voltage element operates as a restraint on the same disc as the overcurrent element, or voltage controlled, where the setting of the overcurrent relay is changed by means of a voltage­operated attracted-armature relay.

OVERLOAD

Overload protection is not generally provided for continuously supervised machines but on large machines resistance thermometers or thermocouples are embedded in the stator winding.  There is some possibility of overload in terms of MVA for, although the governors will restrict MW, the AVR may cause the machine to deliver a disproportionate share of the MVAr.  In cases where overload protection is to be provided this would probably be of the thermal type with a characteristic to match the generator thermal capacity.

Overload and overcurrent relays should not be confused as they perform completely different functions.  An overload relay operates in the hundreds to thousands of seconds range whereas an overcurrent relay operates in the one- to ten-second range.

FAILURE OF PRIME MOVER

In the event of a prime mover failure the generator continues to run but as a synchronous motor and this can cause a dangerous condition in the prime mover.  In a steam turbine the turbulence of the steam In the turbine causes a temperature rise which can quickly reach serious proportions in pass-out sets.  In condensing sets the temperature rise is not as fast and therefore less urgent action is needed.  In engine-driven sets the loss of motive power is likely to be due to mechanical failure and the continued running of the set is likely to cause damage.

The machine, as a synchronous motor, will draw power from the system and it is this reverse power which is detected by the protection.  The power required is usually small, about 10% in case of large turbo-alternators.  The power factor depends on the excitation of the machine and can be quite low and either leading or lagging.  This means that the reverse power relay must respond to a low value of power when the MVAr is high and consequently must be sensitive and have only a small phase-angle error.

A single-element relay is used because the power will be balanced in the three phases.  It is used in conjunction with a time-delay relay to prevent operation during power swings and synchronising.

LOSS OF FIELD

Failure of the field system results in acceleration of the rotor to above synchronous speed where it continues to generate power as an induction generator the flux being provided by a large magnetising component drawn from the system.  This condition can be tolerated for a short time but clearly there will be increased heating of the rotor because of the slip-frequency currents which flow.

Loss of field can be detected by a simple undercurrent relay connected to a shunt in the field circuit.  It must have a setting below minimum field current and a time delay if field forcing is used.  A time-delay relay is also required as the undercurrent relay may respond to the slip-frequency circuit in the field circuit.  This relay would have an instantaneous pick-up and a time-delayed drop off to maintain the circuit to the main time-delayed relay.

The more up-to-date method is to detect the loss of field on the a.c.  side of the generator by comparison of the stator voltage and current.  By either a relay measuring the reactive power (MVAr) which is being imported or by an impedance relay which has a characteristic as shown in Figure 7.4.  As can be seen under normal operation the apparent impedance, as measured by stator voltage and current is well away from the tripping zone.  When there is a loss of field the impedance vector moves to the operation zone.



FIGURE 7.4  -  DETECTION OF GENERATOR FIELD FAILURE

OVERSPEED

The speed is very closely controlled by the governer and is held constant as the generator is in parallel with others in an intercon­nected system.  If the circuit-breaker is tripped the set will begin to accelerate and although the governer is designed to prevent over- speed a further centrifugal switch is arranged to close the fuel valve.

There is still a risk, however, that the fuel valve will not close completely and even a small gap can cause overspeed and so where urgent tripping is not required it is usual to lower the electrical output to about 1% before tripping the circuit-breaker.  A sensitive under-power relay is used to detect when this value is reached.

OVERVOLTAGE

Voltage is generally controlled by a high-speed voltage regulator and therefore overvoltages sbould not occur and overvoltage protec­tion is not generally provided for continuously supervised machines.  On unattended machines an instantaneous relay set at, say 150% is used to cater for defective operation of the voltage regulator.


PROTECTION OF GENERATOR/TRANSFORMER UNITS

Where a generator is connected to the power system by means of a generator transformer it is usual to protect the generator and trans­former as a single unit using biased differential protection.

The current transformer balance is produced in terms of both phase and magnitude, i.e.  in the arrangement shown in Figure 7.5 there is an overall phase change of 30° which is corrected by connecting a set of auxiliary current transformers in delta.  Because of the difference in current transformer ratios the settings of the generator transformer protection has to be somewhat higher than the settings of the generator protection.  Because of this the generator is sometimes protected separately but is also included within the zone of the generator-transformer protection as an extra insurance.  The trans­former is connected directly to the generator and so no harmonic restraint circuit is required in the protection as no switching can occur.  There is a low level of magnetising inrush current following a fault when the voltage is restored from being depressed but this is usually insufficient to unbalance the protection.  Figure 7.5 shows a complete protection system for a generator.



FIGURE 7.5 GENERATOR PROTECTION

No comments:

Post a Comment