Tek-Tips is the largest IT community on the Internet today!
Members share and learn making Tek-Tips Forums the best source of peer-reviewed technical information on the Internet!
-
Congratulations cowski on being selected by the Eng-Tips community for having the most helpful posts in the forums last week. Way to Go!
Frequency relay operation 4
- Thread starter cranky108
- Start date
-
2
- #4
thermionic1
Electrical
Testing the frequency elements on a SEL relay is a bit different that others. Ex if one wanted a definite time of 5 cyc on a 50 element and performed the test, the trip time would very closely match the delay setting. Not so much on SEL frequency elements. I happen to be in front of a SEL-311C with a Doble connected.
Our frequency settings are for event reports and SER only, but it was easy enough to test and report the results.
81D1P(ickup) = 59.3 Hz
81D1d(elay) = 2.0 (cycles)
Test Data (State Simulator)
Prefault Nom V, Nom F for 60cyc. Fault = Nom V, F=59.299 Hz. Start timer at fault state. Trip time = 109.1 mSec. (Doble reports 6.546 cyc (60 cy base))
I wonder if this was considered with the initial settings.
From the SEL application guide AG2018-32 "Setting and Testing Frequency Elements With Time Delay in Cycles" SEL 81 App Guide
"TIME-DELAY CONSIDERATIONS
The power system frequency will not be at nominal value (50 Hz or 60 Hz) when an over- or
underfrequency condition occurs. The relay adjusts processing algorithms to track the system
frequency, and this can make the delay seem shorter or longer than expected. The observed delay
depends on the frequency of the power system during an excursion and whether the frequency
change is applied as a step change, a ramp, or some other function.
Generally, the frequency elements operate on the frequency determined from voltage connected to
the VA-N terminal of the relay. The relay uses a zero-crossing period measurement technique on
the voltage applied to the terminals to determine the power system frequency.
There is an intrinsic delay in the instantaneous frequency elements that depends on the pickup
setting (81DxP), the applied voltage signal, and the conditions prior to a change in signal. System
disturbances that do not cause the undervoltage block element (27B81) to assert can affect voltage
signals and cause the instantaneous frequency elements (81Dx) to assert briefly. It is therefore best
to use a minimum time-delay setting (81DxD) of 5 cycles."
Our frequency settings are for event reports and SER only, but it was easy enough to test and report the results.
81D1P(ickup) = 59.3 Hz
81D1d(elay) = 2.0 (cycles)
Test Data (State Simulator)
Prefault Nom V, Nom F for 60cyc. Fault = Nom V, F=59.299 Hz. Start timer at fault state. Trip time = 109.1 mSec. (Doble reports 6.546 cyc (60 cy base))
I wonder if this was considered with the initial settings.
From the SEL application guide AG2018-32 "Setting and Testing Frequency Elements With Time Delay in Cycles" SEL 81 App Guide
"TIME-DELAY CONSIDERATIONS
The power system frequency will not be at nominal value (50 Hz or 60 Hz) when an over- or
underfrequency condition occurs. The relay adjusts processing algorithms to track the system
frequency, and this can make the delay seem shorter or longer than expected. The observed delay
depends on the frequency of the power system during an excursion and whether the frequency
change is applied as a step change, a ramp, or some other function.
Generally, the frequency elements operate on the frequency determined from voltage connected to
the VA-N terminal of the relay. The relay uses a zero-crossing period measurement technique on
the voltage applied to the terminals to determine the power system frequency.
There is an intrinsic delay in the instantaneous frequency elements that depends on the pickup
setting (81DxP), the applied voltage signal, and the conditions prior to a change in signal. System
disturbances that do not cause the undervoltage block element (27B81) to assert can affect voltage
signals and cause the instantaneous frequency elements (81Dx) to assert briefly. It is therefore best
to use a minimum time-delay setting (81DxD) of 5 cycles."
-
1
- #5
bacon4life
Electrical
Anything relying on zero crossing to determine frequency will get fooled by fault related transients or distorted waveforms. I have been disappointed relay manufacturers do not more clearly state that the 81 element is very likely to misoperate when the time delay is too short. The SEL-300G instruction manual does not appear to give a recommended minimum pickup time. In the 311C the relay will accept a time delay of 2 cycles even though the application guide thermionic1 linked to recommends using at least 5 cycles.
- Thread starter
- #6
I found that the relay setting have too many targets and event reports. The 81 trip did happen, but happened after the generator was off line. I had to go to the plant DCS reporting to find the real tripping of the unit.
That's not to say the under frequency did not trip, just that the breakers were already open.
The frequency operation in the 300G does not have a torque control, which would be nice that I could qualify it with a 52A.
Yes I have experienced 81 relays mis operating on a ground fault, years ago, because of a short time delay. When standards say as quickly as possible, that is misleading, which NERC is finding out.
Thank you all.
That's not to say the under frequency did not trip, just that the breakers were already open.
The frequency operation in the 300G does not have a torque control, which would be nice that I could qualify it with a 52A.
Yes I have experienced 81 relays mis operating on a ground fault, years ago, because of a short time delay. When standards say as quickly as possible, that is misleading, which NERC is finding out.
Thank you all.
How soon after breaker opening? I wonder if the breaker opening caused the zero crossing transient.
The relay should be able to provide an oscillographic record. Suggest polling for the raw data.
A simple AND statement in the trip equation should give you the torque control you want.
The relay should be able to provide an oscillographic record. Suggest polling for the raw data.
A simple AND statement in the trip equation should give you the torque control you want.
Do you know that the UF is based on Voltage or current.
I am not familiar with that relay, but some relays can use current instead of voltage (sometimes a setting preference) and unintentionally using current has caused false operations.
Another possible false trip is when the AC supply to the relay is inadertently connected to an islanded motor/generator supply - there is a bit of generation still happening as it slows down (along with the frequency)
I am not familiar with that relay, but some relays can use current instead of voltage (sometimes a setting preference) and unintentionally using current has caused false operations.
Another possible false trip is when the AC supply to the relay is inadertently connected to an islanded motor/generator supply - there is a bit of generation still happening as it slows down (along with the frequency)
As bacon4life said,"Anything relying on zero crossing to determine frequency will get fooled by fault related transients or distorted waveforms."
Please, read the paper below.
Cheers,
Herivelto S. Bronzeado
Brasília, Brazil
Please, read the paper below.
Cheers,
Herivelto S. Bronzeado
Brasília, Brazil
- Thread starter
- #10
The real event record was overwritten by what happened after the breaker trip. The best I can summarise is the event, the breakers open, the unit over speeds, and trips the boiler. The generator slows down, and the under frequency trips (nothing happens). Exciter trips, and the plant goes black, as the lockout never tripped, so the aux power never transfers.
OOPS. The last relay target is under frequency, which doesn't make since at first.
I know the frequency relays can be fooled, and assumed it was that (my bad).
OOPS. The last relay target is under frequency, which doesn't make since at first.
I know the frequency relays can be fooled, and assumed it was that (my bad).
-
1
- #11
davidbeach
Electrical
Underfrequency relays should only be enabled with normalish voltage and the breaker closed. There's a dedicated voltage element. Breaker status may have to be incorporated in logic. There's also a balance to be reached in event record length vs. triggering criteria so that useful records don't get overwritten before they can be retrieved. Hate it when I can't get an event record I know had to have been written but has since been overwritten.
I’ll see your silver lining and raise you two black clouds. - Protection Operations
I’ll see your silver lining and raise you two black clouds. - Protection Operations
- Thread starter
- #12
As I generally don't do generator settings for the larger units, our consultant that we paid so much for, completed the settings. Our generation engineers are rather young, and are getting a good lesson from this.
I generally only work on T & D.
Life lesson, don't trust anyone who claims to know it all.
I generally only work on T & D.
Life lesson, don't trust anyone who claims to know it all.
- Status
Similar threads
- Question
- Locked
- Question
- Question