You are on page 1of 78

Book 2

Volume 2

Generator
Protection
2005 Protection Seminar

Contents

• Power Plant Horror Stories

6190-118th Avenue North – Largo, Florida 33773


www.beckwithelectric.com
Power Plant
Horror Stories
Presentation

2005 Protection Seminar

6190-118th Avenue North – Largo, Florida 33773


www.beckwithelectric.com
Power Plant
“Horror Stories”
Chuck Mozina
Consultant
Beckwith Electric Co. Inc.
“Horror Stories”

Why I wrote this paper:

ƒ To dispel the myth that nothing exciting ever


happens in power plants
ƒ Many power plant events are not short circuits and
are not easy to analyze (example—loss of field)
ƒ Many event involved human error by less
experienced operators
ƒ With the high volume of IPP plants coming on-line,
commissioning errors are occurring
ƒ “Lessons Learned” are important to avoid repeating
errors in the future
“Horror Stories”

Power Plant Events That The Paper Addresses:

ƒ Multi-Phase Generator Faults


ƒ Stator Ground Current
ƒ Accidental Off-Line Generator Energizing
ƒ Overexcitation
ƒ Loss-of-Field
- ComTrade oscillo analysis
ƒ Generator Breaker Failure
- Breaker Interrupter Flashover
“Horror Stories”

Multi-Phase Generator Faults


“Horror Stories”

Multi-Phase Generator Faults

ƒ Generator trips: generator breaker is


tripped, field is shut down and turbine
tripped.

ƒ Current from system stops, current from


generator continues until stored energy is
dissipated.
“Horror Stories”

I
system

Current I Current Delay


Gen.

0 5 8
Time (sec)
Generator
Breaker Trips

Generator Terminal Fault Current


“Horror Stories”

Multi-Phase Generator Fault Oscillograph

Fault Inception
High Side of
Generator
Breaker
Currents

High Side Breaker Opens

Generator
Neutral
Terminal
Currents
“Horror Stories”

What If Fault is in GSU or UAT?

GSU

G X
UAT F1
R
X
F2
“Horror Stories”

What If Fault is in GSU or UAT?

ƒ If fault is in GSU (F1) or UAT (F2), long


clearing times will result in extensive damage
ƒ A significant number of these transformers
have failed catastrophically with tank rupture
and oil fire
ƒ Low voltage generator breaker significantly
reduces damage
ƒ Sudden pressure relay a good idea because
faults are detected as incipient fault before
high current occurs
“Horror Stories”

“Lessons Learned”

ƒ There is no effective way to quickly “turn off”


generator fault current.
ƒ Long fault decay results in the vast majority of
damage (85%) occurring after tripping.
ƒ Make every effort in design to make sure the
only credible fault in the generator is a ground
fault.
ƒ Consider the value of including a generator
low voltage breaker in your next generator
addition to reduce exposure of GSU and
UAT to long clearing faults.
“Horror Stories”

Stator Ground Faults


“Horror Stories”

Stator Ground Faults

ƒ Generator Tripping: Generator breaker is


tripped, field is shut down and turbine is
tripped.

ƒ Current from system stops when generator


breaker trips, current from generator
continues BUT magnitude is reduced to
generator grounding.
“Horror Stories”
“Horror Stories”

Stator Ground Setting Coordination

ƒ 59N pickup generally set down to 5-6V


pickup to “see” faults near the neutral.

ƒ 59N tripping must be delayed to coordinate


with:
- System ground faults

- Wye-wye generator VT’s fuses


“Horror Stories”

XF1

CTRANS. CTRANS.

VT
X
Fuse
F2

R 59N VN
“Horror Stories”

Line to Ground VT
3 Wire Connection

ƒ Relay VT inputs connected line to line


ƒ This VT connection avoids the need to coordinate
59N with VT fusing
ƒ Can not phase identify stator ground faults
“Horror Stories”

Oscillograph of Field Ground Fault

Generator
Neutral
Voltage
Fault Inception

Breaker Opens

Generator
Phase
Currents

Trip Command
“Horror Stories”

Oscillograph of Field Ground Fault

ƒ Because grounding transformer impedance is


very high compared to generation—full line-to-
neutral voltage appears across transformer.

ƒ Not uncommon for fault to self-extinguish and


re-start because of low ground current.
“Horror Stories”

“Lessons Learned”

ƒ Oscillographic records played a key role in


preventing damaged generators from being
returned to service by confirming relaying
operated properly.

ƒ Oscillographic records can also speed a


generator’s return to service if relaying
operated improperly.
“Horror Stories”

Accidental Off-Line Generator Energizing


“Horror Stories”

Accidental Off-Line Generator Energizing

ƒ Event occurred during commissioning of a


gas turbine installation.

ƒ Commissioning crew was trying to simulate


a 52A contact closer, but jumped the wiring
studs on a terminal block, closing the
breaker.
“Horror Stories”
“Horror Stories”

GSU

Holy Cow! 52G


Aux. / Start-Up
Transformer
I jumpered
the wrong
contact!!

R
“Horror Stories”

Inadvertent Energizing Oscillograph


Generator
Phase
Voltage

Fault Inception

Breaker Opens

Generator
Phase
Currents
“Horror Stories”

Inadvertent Energizing Protection Logic


“Horror Stories”

Inadvertent Energizing Equivalent Circuit

I= Inadvertent Energizing
Current

X2G=Generator Negative
Sequence Reactance

XT =GSU Transformer
Reactance

XS=System Reactance

EG=Generator Terminal
Voltage

ET=GSU High Side Voltage


“Horror Stories”

“Lessons Learned”

ƒ Despite the design of modern interlock


schemes to prevent it—inadvertent energizing
events continue to happen.

ƒ Dedicated inadvertent energizing protection is


needed even on new gas turbine plants with
generator low voltage breakers.

ƒ Inadvertent energizing schemes need to be in-


service when the generator is out of service.
“Horror Stories”

Overexcitation
“Horror Stories”

Overexcitation

ƒ V/Hz relaying used to protect generator and


power plant transformers from excessive
magnetic flux.
ƒ Excess flux level overheats generator and
transformer core steel.
ƒ Core flux proportional to voltage and
inversely proportional to frequency—thus
V/Hz protection.
ƒ Most V/Hz events occur when generator is
off-line prior to synchronizing.
“Horror Stories”

Overexcitation/Volts per Hertz (24)

GENERATOR
Voltage V
TRANSFORMER ≈
Freq. Hz
EXCITATION

GENERATOR LIMITS (ANSI C 50.13)


Full Load V/Hz = 1.05 pu
No Load V/Hz = 1.05 pu
TRANSFORMER LIMITS (ANSI C57.12)
Full Load V/Hz = 1.05 pu (HV Terminals of GSU)
No Load V/Hz = 1.10 pu (HV Terminals of GSU)
“Horror Stories”

Overexcitation/Volts per Hertz (24)

PHYSICAL INSIGHTS
ƒ As voltage rises above rating leakage flux increases
ƒ Leakage flux induces current in transformer support
structure causing rapid localized heating
“Horror Stories”

Overexcitation/Volts per Hertz (24)


TYPICAL CURVES
“Horror Stories”

Overexcitation/Volts per Hertz (24)

CAUSES OF V/HZ PROBLEMS


ƒ Generator voltage regulator problems
- operating error during off-line manual regulator
operation
- control failure
- loss of VT regulator supply voltage
ƒ System problems
- unit load rejection: full load, partial rejection
- power system islanding during major
disturbances
“Horror Stories”

V/Hz Event #1
“Horror Stories”

VA VB VC IA IB IC
Overview

GEN. BREAKER OPEN


TRIP
ALARM
ALARM
“Horror Stories”

VA & I A

GEN. BREAKER OPEN TRIP


ALARM
ALARM
“Horror Stories”

IA

GEN. BREAKER OPEN


TRIP
ALARM
ALARM
“Horror Stories”

Event #2

Open Prior to Synchronizing

Open VT

AVR V/Hz Full Voltage


Voltmeter

R
“Horror Stories”

“Lessons Learned”

ƒ Most V/Hz events occur when the generator


is off-line.

ƒ VT open circuit to the AVR is a frequent


condition that causes V/HZ events.

ƒ V/HZ protection should be in a different VT


circuit than the AVR to prevent a single
open VT from causing a V/Hz condition and
at the same time disabling protection.
“Horror Stories”

Loss-of-Field (40)
“Horror Stories”

Loss-of-Field (40)

ƒ Detriments
- Generator
ƒ Synchronous generator becomes induction
ƒ Slip induced eddy currents heat rotor surface
ƒ High reactive current drawn by generator
overloads stator
“Horror Stories”

Loss-of-Field (40)

ƒ Detriments
- Power system
ƒ Loss of reactive support
ƒ Creates a reactive drain
ƒ Can trigger system/area voltage collapse
“Horror Stories”

Generator Capability Curve


“Horror Stories”

Modern Loss-of-Field Using Two-Zone Off-


Set mho Method
+X

-X´d Heavy Load Light Load


2
-R +R

Xd
1.0 pu
{ Impedance Trajectory
on Loss-of-Field

Machine Capability
Minimum Exciter Limit

-X
“Horror Stories”

Loss-of-Field (40)

ƒ Causes
- Field open circuit

- Field short circuit

- Accidental tripping of field breaker

- Regulator control failure

- Loss of main exciter


“Horror Stories”

Loss-of-Field (40) - What the operator


wanted to do
Utility Tie
ƒ Desired sequence Turbine
of events: T
Field Breaker
- Turbine was tripped
due to oil valve
solenoid failure GEN
42 MVA
- Sequential tripping
scheme failed to shut Incoming
Gen.
down unit Breaker A B Breaker

- Operator wanted to BUS


trip generator breaker TIE
Tie Bus
A and then field
breaker
“Horror Stories”

Loss-of-Field (40) - What the operator


actually did
Utility Tie
Actual sequence of Turbine

events: T
Field Breaker
ƒ Operator mistakenly
opened in-coming GEN
42 MVA
Breaker B
„ Then tripped Field Gen.
A B
Incoming
Breaker
Breaker
Breaker
BUS
„ RESULT—Generator TIE
Tie Bus
connected to system
with no field
“Horror Stories”

Loss-of-Field (40) Oscillograph

Turbine Valves Closed


Field Breaker Trip

Relay Trip
“Horror Stories”

X ∞
R-X Diagram Analysis Impedance Locus
Using ComTrade From ComTrade Simulation

26.1(T=0)

R
2.0Ω

27.7Ω

trip point


“Horror Stories”

“Lessons Learned”

ƒ Operating errors are a significant cause of loss-of-


field events.
ƒ Relay oscillographs provide unbiased evidence of
what actually happened.
ƒ Use of programmable inputs to monitor turbine
values and generator breaker(s) positions help
document sequence of events.
ƒ ComTrade provides a handy tool to verify loss-of-
field relay operation.
“Horror Stories”

Generator Breaker Failure (Open Breaker


Flashover)
“Horror Stories”

Open Generator Breaker Flashover on


Mexican Utility System
Voltage Across
Open Breaker Prior
to Synchronizing
Vsystem
Vgen
Vge
n

Note: Tripping of generator Vgen Vge


breaker does not arrest the n
Vge
event. Need to activate n
breaker failure.
“Horror Stories”

Head Flashover Equivalent Circuit


“Horror Stories”

Basic Generator Breaker Failure Logic

One-Line Diagram

Generator Breaker Failure Logic

Key item – current detector (CD) must be set below flashover


current. In Mexico, current detector was set above
flashover current.
Result – Complete Failure of Rotor
“Horror Stories”

Modified Breaker Failure For Flashover


Protection

One-Line Diagram

Generator Breaker Failure Logic

ƒ Current Detector (CD) must still be set below


flashover current.
ƒ Use of 50N relay speeds up B.F.I.– Mexican utility
made this modification.
“Horror Stories”
Conclusions

Conclusions
ƒ Generator events are not as rare as some people believe
ƒ They are very disruptive and costly to a utility or IPP
owner. These events immediately cost the generator
owner $
ƒ Oscillographs and sequence-of-event data from digital
relays play a key role in determining what really
happened
ƒ This greatly helps to keep damaged generators off-line
and in returning undamaged machines to service
ƒ Many power plant events involve human error—may be
the result of downsizing
ƒ We learn from our experiences—or we are DOOMED to
repeat them
“Horror Stories”

The End

Power Plant
“Horror Stories”
Chuck Mozina
Consultant
Beckwith Electric Co. Inc.
Power Plant
Horror Stories
Paper

2005 Protection Seminar

6190-118th Avenue North – Largo, Florida 33773


www.beckwithelectric.com
Power Plant “Horror Stories”
Charles J. Mozina
Beckwith Electric Co., Inc.
6190-118th Avenue North
Largo, FL 33773-3724 U.S.A.
(727) 544-2326

Abstract: Contrary to popular belief, generators can fail—not only from short circuits—but more fre-
quently from abnormal electrical conditions such as overexcitation, overvoltage, loss-of-field, unbal-
anced currents, and abnormal frequency. When subjected to these abnormal conditions, damage or
complete failure of the generator can occur within seconds. In many cases, these failures can be pre-
vented by proper generator protection. This paper relates a number of “horror stories” within the power
plant in the hopes that the “lessons learned” will help others to avoid the cases described.

Introduction
Generators are the most expensive piece of equipment in a power system. The cost of a major generator
failure to a utility or IPP (Independent Power Producer) owner is not only the cost of repair or replace-
ment of the damaged machine, but also the substantial cost of purchasing replacement power when the
unit is out of service. An alert and skillful operator, at manned locations, can sometimes avoid remov-
ing a generator from service by correcting an abnormal condition. In the vast majority of cases, how-
ever, the event will occur too rapidly for the operator to react and automatic detection and isolation is
required. Operators have also been known to make errors and create abnormal conditions where trip-
ping to avoid damage is required. Inadvertent energization and overexcitation are examples of such
events.

Several power plant events within the last two to three years substantiate the premise that generators
can, and do, sustain internal short circuits and abnormal operating conditions that require tripping. The
following in-service events are described:

• Multi-phase generator faults


• Stator ground faults
• Accidental off-line generator energizings
• Overexcitation
• Loss-of-field
• Generator breaker failure (breaker flashover)

In many cases, human error caused or contributed to the event. These events were captured on oscillo-
graphs. This paper highlights the subtlety of analyzing non-fault events such as loss-of-field using
COMTRADE format to convert current and voltage to R-X quantities to verify proper relay operation.
The lessons learned in each event are also highlighted.
Multi-Phase Generator Faults
When a generator multi-phase fault is detected by generator differential relaying, it is separated from
the power system by tripping the generator breaker, field breaker and prime mover. The system contri-
bution to the fault will immediately be removed when the generator breaker trips as illustrated in Fig. 1.
The generator current, however, will continue to flow after the trip. The generator short circuit current
cannot be “turned off” instantaneously because of the stored energy in the rotating machine. This flow
of damaging generator fault current will continue for several seconds after the generator has been
tripped, making generator faults extremely damaging. Generator terminal leads are usually isolated
through isophase bus construction to minimize multi-phase terminal faults.
GSU
IGEN (a,b,c) ISYSTEM (A,B,C)
Power
G X System
R
Multi-Phase Fault

VN
I
system

Current I Current Delay


Gen.

0 5 8
Time (sec)
Generator
Breaker Trips

Fig. 1 Generator Terminal Fault Current

Fig. 2 is an oscillograph of a three-phase fault which occurred on a gas turbine when a connector failed
at the generator lead connection to the generator breaker. The fault started as a line-to-ground fault, but
after five cycles, it evolved into a three-phase fault. The system currents (IA, IB, IC) were interrupted
when the generator breaker was opened by differential (87G) relaying in about three cycles. The gen-
erator-side current (Ia, Ib, Ic) continued to flow after the unit was shut down. The oscillograph was
programmed to cut off six cycles after tripping, thereby preventing the display of the total length of
fault current flow which is estimated to have continued for eight seconds after tripping.

This extended flow of fault current is the reason that internal multi-phase generator faults typically damage
the unit to the point where it cannot be economically repaired. There is no means of “turning off” the
generator current. This long decay time results in the vast majority (about 85%) of the damage occurring
after tripping. This is why every effort is made in generator and generator terminal design to make the only
credible fault a ground fault. The generator is then grounded so as to substantially reduce ground current to
minimize damage. If the fault is in the GSU transformer and the generator installation has no low-voltage
breaker, the long fault current decay can substantially damage the transformer. A significant number of these
transformers have failed catastrophically with tank ruptures and oil fires.
Breaker Open
Relay Trip
Fig. 2 Multi-Phase Generator Fault Oscillograph

Stator Ground Faults


The method of stator grounding used in a generator installation determines the generator’s performance
during ground fault conditions. If the generator is solidly grounded (not usually the case), it will deliver a
very high current to a SLG (single-line-to-ground) fault at its terminals, accompanied by a 58% reduction in
the phase-to-phase voltages involving the faulted phase and a modest neutral voltage shift. If the generator is
ungrounded (also not usually the case), it will deliver a negligible amount of current to a bolted SLG fault at
its terminals, accompanied by no reduction in the phase-to-phase terminal voltages and a full neutral voltage
shift. These represent the extremes in generator grounding with normal practice falling predictably in between.

The high magnitude of fault current which results from solidly grounding a generator is unacceptable be-
cause of the fault damage it can cause. Shutting down the generator through tripping the generator breaker,
field, and prime mover does not cause the fault current to immediately go to zero. The flux trapped in the
field will result in the fault current slowly decaying over several seconds after the generator is tripped—
substantially exacerbating damage. On the other hand, operating an ungrounded generator provides negli-
gible fault current, but the line-to-ground voltages on the unfaulted phases can rise during arcing type faults
to dangerously high levels which could cause the failure of generation insulation. As a result, stator windings
on major generators are grounded in a manner that will reduce fault current and overvoltages yet provide a
means of detecting the ground fault condition quickly enough to prevent iron burning.
Almost all large generators that are unit-connected are high-impedance grounded. High-impedance
generator neutral grounding utilizes a distribution transformer with a primary voltage rating greater
than or equal to the line-to-neutral voltage rating of the generator and a secondary rating of 120 V or
240 V. The distribution transformer should have sufficient overvoltage capability so that it does not
saturate on SLG faults with the machine operating at 105% of rated voltage. The secondary resistor is
usually selected so that for a SLG fault at the terminals of the generator, the power dissipated in the
resistor is approximately equal to the reactive volt-amperes in the zero-sequence capacitive reactance
of the generator windings, its leads, and the windings of any transformer(s) connected to the generator
terminals. Using this grounding method, a SLG fault is generally limited to 3-5 primary amperes. As a
result, this level of fault current is not sufficient to operate generator differential relays.

Fig. 3 illustrates a typical unit-connected high-impedance grounded generator. The most widely used
protective scheme in high-impedance grounded systems is a time-delayed overvoltage relay (59N)
connected across the grounding resistor to sense zero-sequence voltages as shown in Fig. 3. The relay
used for this function is designed to be sensitive to fundamental frequency voltage and insensitive to
third-harmonic and other zero-sequence harmonic voltages that are present at the generator neutral.

VT

VN
R 59N

Fig. 3 Unit-Connected High Impedance-Grounded Generator

Since the grounding impedance is large compared to the generator impedance and other impedances in the
circuit, the full phase-to-neutral voltage will be impressed across the grounding device for a phase-to-ground
fault at the generator terminals. The voltage at the relay is a function of the distribution transformer ratio and
the location of the fault. The voltage will be a maximum for a terminal fault and decreases in magnitude as
the fault location moves from the generator terminals toward the neutral. Typically, the overvoltage relay has
a minimum pickup setting of approximately 5 V. With this setting and typical distribution transformer ratios,
this scheme is capable of detecting faults to within approximately 5% of the stator neutral. Third harmonic
schemes (not described in this paper) are typically used to detect faults near the generator neutral.
The time setting for the overvoltage relay is selected to provide coordination with other system protec-
tive devices. The voltage relay should be coordinated with the transmission system relaying for system
ground faults. System phase-to-ground faults will induce zero-sequence voltages at the generator neu-
tral due to capacitive coupling between the windings of the GSU transformer. This induced voltage will
appear on the secondary of the grounding transformer and can cause operation of the zero-sequence
voltage relay. When grounded wye-grounded-wye VT’s are connected at the machine terminals, the
neutral ground overvoltage relay should be coordinated with VT transformer fuses to prevent tripping
the generator for VT secondary ground faults.

Fig. 4 shows an oscillograph for a stator ground fault that occurred in a large unit-connected generator
in the southeastern U.S.. Note that because of the very low level of ground fault current, it is not
uncommon for the fault to self-extinguish and then re-establish itself. Also, the normal third harmonic
voltage across the neutral resistor shifts to fundamental frequency when a ground fault occurs. By
measuring the magnitude of the neutral voltage and comparing it to the calculated value for a terminal
fault, you can determine the approximate fault location in relationship to the generator terminal.

The oscillograph shown in Fig. 4 played a key role in preventing a damaged generator from being
returned to service by the power plant manager. When the tripping occurred, the new digital relay had
only been in service for a few months. The generator stator windings were meggered, but the voltage
produced was not adequate to break down the ground. The plant manager was ready to return the unit to

Fig. 4 Stator Ground Fault Oscillograph Relay Trip


service. The oscillograph provided documented evidence that the ground fault did, in fact, occur. Based
on the oscillographic data, the decision was made to bring a high voltage Hi-Pot Test set from another
plant. The resulting test uncovered the ground fault which was caused by a cracked generator terminal
bushing. The bushing was replaced and the unit was returned to service.

Accidental Off-Line Generator Energizing


Inadvertent or accidental energizing of synchronous generators has been a particular problem within the industry in
recent years. A number of machines have been damaged or, in some cases, completely destroyed when they were
accidentally energized while off-line. The frequency of these occurrences has prompted generator manufacturers
to recommend that the problem be addressed through dedicated protective relay schemes. Operating errors, con-
trol circuit malfunctions, or a combination of these causes, have resulted in generators becoming accidentally
energized while off-line. In modern gas turbine applications, the major cause of inadvertent energization of genera-
tors has been by closing the generator breaker through the mechanical close/trip control at the breaker itself,
thereby defeating the electrical interlocks. Fig. 5 illustrates a typical gas turbine one-line diagram configuration.

During the commissioning of a new gas turbine in southeast Georgia, the commissioning crew was trying to
simulate a generator breaker 52A contact closure by jumping the contact at a terminal block. The wrong
terminals were inadvertently jumped, resulting in the generator breaker closing onto a “dead” machine. Fig.
6 shows the oscillograph of the event.

GSU

Generator Aux. / Start-Up


52G
Breaker Transformer

Fig. 5 Gas Turbine Inadvertent Energizing


Due to the severe limitation of conventional generator relaying to detect inadvertent energizing, dedicated
protection schemes have been developed and installed. Unlike conventional protection schemes, which
provide protection when equipment is in service, these schemes provide protection when equipment is out of
service. One method widely used to detect inadvertent energizing is the voltage-supervised overcurrent
scheme shown in Fig. 7. An undervoltage element with adjustable pickup and dropout time delays supervise
an instantaneous overcurrent relay.
Breaker Closed

Relay Trip
Fig. 6 Inadvertent Energizing Oscillograph

The undervoltage detector automatically arms the overcurrent tripping when the generator is taken off-line.
The undervoltage detector will disable or disarm the overcurrent relay when the machine is returned to
service. Great care should be taken when implementing this protection, so that the DC tripping power is not
removed when the generator is off-line.

When an off-line generator is energized while on turning gear or coasting to a stop, it behaves as an
induction motor and can be damaged within a few seconds. During three-phase energization at a stand-
still, a rotating flux at synchronous frequency is induced in the generator rotor. The resulting rotor
current is forced into paths in the rotor body, similar to those rotor current paths for negative sequence
stator currents during generator single-phasing. Rapid rotor heating, and damage to the rotor, will occur.

50
Overcurrent
I > P. U .

Pickup
AND
Delay Output
27 Contact
Undervoltage*
V<P.U. Dropout
Delay

Fig. 7 Inadvertent Energizing Scheme


The machine impedance during this high-slip interval is equivalent to the generator negative-sequence
reactance. Fig. 8 shows a simplified equivalent circuit that can be used to calculate the current and
voltage associated with three-phase inadvertent energizing.

I I= Inadvertent Energizing Current

X2G=Generator Negative Sequence Reactance


XT
X1S
X2G XT =GSU Transformer Reactance
E ES
EG T Equivalent XS=System Reactance
System
Voltage
EG=Generator Terminal Voltage

ET=GSU High Side Voltage

Fig. 8 Inadvertent Energizing Equivalent Circuit

Overexcitation
Overexcitation, or V/Hz, relaying is used to protect generators from excessive magnetic flux density
levels. High flux density levels result in an overexcitation of the generator. At high-flux levels, the
magnetic iron paths designed to carry the normal flux saturate, and flux begins to flow in leakage paths
not designed to carry it. These resulting fields are proportional to voltage and inversely proportional to
frequency. Hence, high flux density levels (and overexcitation) will result from overvoltage, underfre-
quency, or a combination of both. ANSI/IEEE Standards C50.13 and C57.12 for generators and trans-
formers have established the following V/HZ limits for continuous operation:

Generators 1.05 p.u. (generator base)


Transformers 1.05 p.u.; (on transformer secondary base) at rated load, 0.8 pf or
greater; 1.1 p.u. (transformer base) at no load

These limits apply, unless otherwise stated by the equipment manufacturer. When these V/Hz ratios are
exceeded, saturation of the magnetic core of the generator or connected transformer(s) will occur.
Damage can occur within seconds. It is the general practice to provide V/Hz relaying to protect genera-
tors and transformers from these excessive magnetic flux density levels. This protection is typically
independent of V/Hz control on the excitation system.

Damage due to excessive V/Hz operation most frequently occurs when the unit is off-line, prior to
synchronization. The potential for overexcitation of the generator dramatically increases if operators manually
prepare the unit for synchronization. This is particularly true if the overexcitation alarms are inadequate, or if
the VT has an open circuit due to an improper connection. This was the case for a large unit-connected
system generator in the midwestern United States. The one-line diagram is shown in Fig. 9.

The unit was being returned from service after an outage where work had been done on the AVR
(Automatic Voltage Regulator). To isolate the regulators, the VT fuses had been pulled, but they were
not re-installed after the work was completed. When the unit was brought on-line prior to synchronizing,
Open Generator Breaker

GSU

VT Unit
Auxillary
Open
VT
AVR
R

Fig. 9 Overexcitation One-Line Diagram

the operator transferred from manual to automatic AVR control. The AVR, sensing low voltage, went to
full boost increasing field current to the maximum level in attempting to raise the generator terminal
voltage. Fig. 10 is an oscillograph of the event. The voltage rose to 120% of normal and current to 30%
of generator full load before the trip.

Trip
Alarm
Alarm
Fig. 10 Overexcitation Oscillograph
Fortunately, both generators and transformers can sustain V/Hz levels above their continuous capability for
a substantial time (seconds) since it takes time to heat the iron core of both devices. Therefore, no damage
was done to the generator. Operating procedures were revised to check VT continuity to the AVR prior
to switching to automatic operation.

In another case, in a large industrial plant, a machine was completely destroyed when the VT open
circuit disabled both the AVR and protection. Fig. 11 illustrates the one-line diagram. A “stuck” volt
meter provided the operator with a false voltage reading. As he increased field current to raise terminal
voltage, he overexcited the generator until it failed. The lesson here is that the protection should be in
a separate VT circuit from the AVR control and voltmeter.

Open Prior to Synchronizing

Open VT

AVR V/Hz V

Fig. 11 Overexcitation - stuck voltmeter

Loss-of-Field
Partial or total loss-of-field on a synchronous generator is detrimental to both the generator and the power
system to which it is connected. The condition must be quickly detected and the generator isolated from the
system to avoid generator damage. When the generator loses its field, it operates as an induction generator,
causing rotor heating. A loss-of-field condition which is not detected can have a devasting impact on the
power system by causing a loss of reactive power support as well as creating a substantial power drain. If not
detected quickly on large generators, this condition can trigger a system voltage collapse.

If the excitation current is reduced or lost, the generator absorbs reactive power from the power system, rather
than supplying it, and it operates in the underexcited region of the capability curve. If a total loss of field occurs
and the system can supply sufficient reactive power without a large terminal voltage drop, the generator will run
as an induction generator, otherwise synchronism will be lost. The change from normal overexcited operation
to underexcited operation upon loss-of-field is not instantaneous, but occurs over a time period (generally
cycles) depending on the generator’s output level and connected system capability. The generator’s capability
curve (Fig. 12) outlines the generator’s operating limits.
Reactive Power Rotor
into System Winding
Limited
+ MW
MVAR
G System
MVARS
Overexcited Stator Normal Overexcited
Winding Operation
Limited
+MW
0 Real Power
Into System

Minimum MW
Underexcited Excitation
Limiter G System
MVARS
MVAR Underexcited Operation
Stator End
Reactive Power Iron Limited
into Generator

Fig. 12 Generator Capability Curve

The most widely applied method for detecting a generator loss-of-field is the use of distance relays to
sense the variation of impedance as viewed from the generator terminals. It has been shown that when a
generator loses its excitation while operating at various levels of loading, the variation of impedance as
viewed from the machine terminals will have the characteristics shown on the R-X diagram in Fig. 13.

+X

-X´d Heavy Load Light Load

{
2
-R +R

Xd
1.0 pu
{ Impedance Trajectory
on Loss-of-Field

Machine Capability
Minimum Exciter Limit
-X

Fig. 13 Modern Loss of Field Using Two-Zone Off-Set mho Method

The loss-of-field relay measures the impedance as viewed from the machine terminals, and it operates
when the impedance falls inside the circular characteristic. The relay is offset from the original by one-
half of the direct axis transient reactance (X´d/2) to prevent misoperation during system disturbances
and other fault conditions. The diameter of the circle is adjusted to be equal to the direct axis synchro-
nous generator reactance. A time delay is used to provide security against false tripping for stable power
swings. This time delay increases the operating time of the relay, which means that the MVArs drawn
by the generator persist for a longer time, making the system more susceptible to severe voltage dips.
Many users have upgraded to modern two-zone mho relays to enhance protection. This scheme is
shown in Fig. 13. The inner small mho circle is set to trip with only a few cycles of delay and is within
the impedance locus trajectory for most loss-of-field events. The fast operation of the inner mho unit
quickly detects a loss-of-field condition.

The locus of the impedance trajectory for a loss-of-field condition depends on the value of system
impedance. For machines connected with system impedances approximately less than 20% of the gen-
erator X´d , the impedance locus will take a direct path to the final point. With higher system imped-
ances, the trajectory will spiral to the final point. If the machine is fully loaded prior to the loss-of-
excitation condition, the machine at the final impedance point will be operating as an induction genera-
tor, with a slip of 2 to 5% above normal. The machine will also start receiving reactive power from the
system while supplying reduced real power.

Complete loss-of-excitation occurs when the DC source of the generator field is interrrupted. The loss of
excitation can be caused by such incidents as an open circuit in the field, a short circuit in the field, accidental
tripping of the field breaker, regulator control system failure, loss-of-field to the main exciter, or loss of ac
supply to the excitation system. The accidental trip of the field breaker can occur in the “heat of battle” when
the unit operator is trying to react to a number of events occurring in rapid succession. Oscillography and
sequence-event recorders are often used to “sort out” the event and uncover what happened. This was the
case during a cascading event at a large paper mill. The one-line diagram is shown in Fig. 14.

Utility Tie
Turbine

Field Breaker
T

GEN
42MVA

Gen. Incoming
Breaker A B Breaker

BUS
TIE
Tie Bus
Fig. 14 Paper Mill One-Line Diagram

The event occurred shortly after start-up with a significant load on the unit when the turbine oil trip
device failed— dumping the oil header and closing the main stop valve. The oil trip solenoid linkage
had worn badly, causing failure. Reverse power relaying picked up, but was unable to trip. Upon
realizing the main generator breaker had not tripped, the operator mistakenly tripped the utility tie
(breaker B in Fig. 14) instead. He followed by tripping the field breaker about seven seconds later.
Detecting the loss-of-field condition, the loss of field protection tripped the generator just in time to
avert a mill shutdown from undervoltage from the reactive power upset. Sequence of events and oscil-
lography captured by the digital relays verify the following:
1. The relaying verified the repeated reverse power (32) pick-ups but the 32 relay could not trip the
main and field breakers through lockout relays due to a wiring error. For about two minutes after the
oil trip linkage failure, the generator motored with the field, supplying up to 25 MVArs to support
bus voltage after the accidental loss of the incoming utility tie. The current can be seen to be 90
degrees lagging the voltage in the oscillograph in Fig. 15.

2. When the field breaker opened, reactive power flow reversed in 18 cycles and the generator drew in
leading current rising to 800 amps from the system to re-establish field as shown in Fig. 15. The unit
was now operating without field as an induction machine and placing a large reactive load on the
mill system, with the bus voltage sagging to about 11 kV on a 13.8 Kv system.

Turbine Values Closed

Field Breaker Trip Relay Trip


Fig. 15 Loss of Field with Turbine Tripped

3. Sensing the loss-of-field condition, the loss-of-field relaying tripped the unit about two seconds
after the field breaker had been opened. Separate bus tie relay data verified the tie was in an
overload condition and the overcurrent relay was timing to trip the bus for a total bus shutdown.

4. Initially, the operator believed he had followed normal operating procedure and that the utility tie had
opened on overload from the loss of the generator. The relay data provided unbiased evidence to
show that he had misoperated and tripped the utility tie breaker.
X

26.06(T=0)
C Impedance Locus
From ComTrade Simulation
R
2.0Ω

-27.656
trip point

Fig. 16 R-X Diagram of Loss of Field Event with Turbine Tripped


To further analyze this event, the oscillographic data was converted to COMTRADE format to allow it
to be conveniently plotted on the loss-of-field R-X diagram. The R-X diagram plot is shown in Fig. 16.

When a loss-of-field event occurs on a machine prior to the turbine tripping, the oscillograph looks
much different. The event depicted on the oscillograph in Fig. 17 occurred on a large steam generator.
Again, an operating error resulted in the field breaker being tripped with the generator breaker still
closed. The machine tripped by loss-of-field relaying. A key in analyzing loss-of-field events is that the
current will lead the voltage resulting in VARS being taken into the generator. The use of COMTRADE
to “play” the event back to the relay provides an excellent tool in analyzing loss-of-field events.

Fig. 17 Loss of Field Prior to Turbine Trip


Generator Breaker Failures
A unique form of generator high-voltage breaker failure is the case where the breaker contact flashover
energizes the generator. Fig. 18 illustrates such a flashover case. This type of flashover has occurred on
a number of EHV air-blast breakers where SF6 pressure was lost prior to the unit being synchronized to
the system. With the loss of dialectic in the interruptor, the contact flashed across when a 2.0 per unit
voltage is placed across the interruptor in each slip cycle. Because the 2.0 per unit voltage only occurs
on generator breakers, this type of flashover is more likely on these breakers. Fig. 19 shows the equiva-
lent circuit that can be used to calculate the resulting currents.

Fig. 18 Open Breaker Flashover

X10′X2G′X0G = Generator Positive, Negative and Zero Sequence Reactances


X1T′ X2T′ X0T = Unit Step-Up Transformer Positive, Negative and Zero Sequence Reac-
tances
X1S′ X2S′ X0s = System Equivalent Positive, Negative, and Zero Sequence Rectances
Eg = Generator Voltage
ES = System Voltage
I1′ I2′ I0 = Positive, Negative and Zero Sequence Currents
Fig. 19 Headflashover Equivalent Circuits
A breaker generator flashover event recently occurred on a large steam generator in Mexico. The one-line
configuration is shown in Fig. 20a. For the flashover of a generator high voltage breaker pole, re-tripping of
the breaker will not de-energize the machine because the breaker is already open. The initiation of breaker
failure is required to trip additional local, and possibly remote, breakers to de-energize the generator.
52/a

TRIPS
OR BACKUP
CD AND TIMER
BREAKERS
AND
A BFI UNIT

G
Power PROTECTIVE
RELAYS 52/a - CIRCUIT BREAKER
System AUXILIARY CONTACT

B CD - CURRENT DETECTOR

CD BFI - BREAKER FAILURE INITIATE

a) One-Line Diagram b) Generator Breaker Failure Logic


Fig. 20 Breaker Flashover and Breaker Failure Logic

Unbalanced currents associated with breaker head flashover will generally cause the generator nega-
tive-sequence relay to operate. This will initiate tripping of the generator breaker(s) (which are already
open), shutting down the generator and providing the BFI signal (Breaker Failure Initiate) to the breaker
failure. The Breaker Failure Logic (Fig. 20b) breaker failure will be initiated only if the breaker failure
current detectors are set with sufficient sensitivity to detect the flashover current. The equivalent circuit
in Fig. 19 provided the circuitry to calculate the flashover current. In the case of the unit in Mexico, the
current detectors were set higher than the flashover current and breaker failure was not initiated. The
result was the catastrophic failure of the generator. The key consideration for setting the breaker failure
current detector is to set it below the level of the current expected for breaker-pole flashovers. This was
an expensive lesson learned by this Mexican utility.

One approach used to speed the detection of a breaker flashover is to modify the breaker failure scheme as
shown in Fig. 21. An instantaneous overcurrent relay (50N) is connected in the neutral of the generator step-up
transformer and is set to respond to an EHV breaker pole flashover current. The relay output is supervised by the
generator breaker “52B” contact providing an additional start to the breaker failure scheme. When the generator
breaker is open and one or two poles of the breaker flash over, the resulting transformer neutral current is
detected by the 50N relay without the delay that would be associated with a negative sequence relay. The current
detectors (CD) associated with the generator breaker scheme must be set with sufficient sensitivity to detect
this flashover condition. This scheme is used by a number of utilities and is also described in ANSI/IEEE
C37.102 (Guide for the Protection of Synchronous Generators). The Mexican utility that sustained the gen-
erator failure retrofitted their breaker failure with this scheme.
52/a

TRIPS
OR BACKUP
CD AND TIMER
BREAKERS
S1 A AND
BFI UNIT
GEN
PROTECTIVE OR
RELAYS 52a, 52b - CIRCUIT BREAKER
B AUXILIARY CONTACT

CD - CURRENT DETECTOR
C D 50N CD
50N 52b 50N - ISTANTANEOUS
Modification OVERCURRENT RELAY
BFI - BREAKER FAILURE INITIATE

a) One-Line Diagram b) Generator Breaker Failure Logic

Fig. 21 Modified Breaker Flashover and BreakerFailure Logic

Conclusion
Major generator tripping events are not as rare as many people believe. Should they occur, such events
can be very disruptive and costly to utility power production. Delays in determining the cause of the
disruption and in assessing equipment damage can add hours to reenergizing and returning to normal
operations. It is critical to assess the integrity of the generator and perform any testing necessary before
reconnecting to the system. It is critical to have good sequence-of-events and oscillographic data to
review protective relay performance, assess fault levels and clear the machine of any damage. To pro-
ceed otherwise, risks restarting the unit with possible increased damage and extensive downtime.

The loss of a generator immediately costs the utility money for lost production and there is great pressure
placed on those involved to determine if the generator is damaged, determine what generator testing
needs to be done and return the machine to service as quickly as possible. Oscillographs and sequence-
of-events data from a digital relay have greatly aided in this effort.

References
1. ANSI/IEEE C37.102-1995. “Guide for AC Generator Protection”.
2. IEEE 95TP102. “IEEE Tutorial on the Protection of Synchronous Generators”.
3. C.J. Mozina “Upgrading Generator Protection Using Digital Technology.” Presented at the 1995
Georgia Tech Relay Conference.
4. C.J. Mozina “Advanced Applications of Multifunction Digital Generator Protection.” Presented
at the 2001 Georgia Tech Relay Conference.
5. ANSI/IEEE C57.12.00 “Standard General Requirements for Liquid-Immersed Distribution Power
and Regulating Transformers”.
6. Powell, L.J. “The Impact of System Grounding Practices on Generator Fault Damage,” IEEE Trans
actions on Industrial Applications, Vol. 37, Jan./Feb. 2001, pp. 218-222.
7. Clarke, E. “Circuit Analysis of A-C Power Systems”, Vol. I; Copyright 1943, General Electric Company.
000
2
1:
d
re
0
te
90

g is
Re
IS O

BECKWITH ELECTRIC CO., INC.


6190 - 118th Avenue North • Largo, Florida 33773-3724 U.S.A.
Phone: (727) 544-2326 • Fax: (727) 546-0121
E-mail: marketing@beckwithelectric.com
Web Site: www.beckwithelectric.com

© 2005 Beckwith Electric Co.


Printed in U.S.A.

You might also like