veritas
Electrical
- Oct 30, 2003
- 467
I have another generator protection question which came up recently. I am looking at a 600MW steam turbine set generating at 22kV, neutral is high z gounded so that EF current < 10A.
The existing EF consists of:
1. 0.5A primary pickup, 5s time delay fed from CT in neutral. (Main 1)
2. 10V, 0.5s neutral overvoltage (NOV) fed from the neutral earthing trfr (essential a single phase distribution trfr). (Main 2)
Am now applying:
1. 100% stator EF in the SEL-700G using 3rd harmonic UV & NOV as well as retaining the 0.5A, 5s NEF relay (Main 1 protection).
2. 100% stator EF based on 20Hz injection (Main 2) as well as retaining the NOV function.
There has been quite some debate about what the trip times should be!
First thing is that any EF operation will result in Turbine Trip only. The GCB will then trip on RP in about 2s (soft landing) so one has to factor in around 2s additional delay to the trip time.
My view is a 2s trip for all EF functions which means fault will be removed in around 4s + CB time. Is there any point in a longer delay?
However, I am also keen to have a NOV element set to 80% of full displacement voltage which means it will detect faults at the machine terminals or 22kV network. Trip time now is around 0.5s. My concern here is for the voltage rise on the healthy phases with EF's. I wish to avoid a flashover with it's serious consequences.
Please also note that the two auxiliary trfrs as well as the GSU all have 22kV delta windings so the gen EF does not see EHV system faults.
I am keen to know what works for others.
Thanks in advance.
The existing EF consists of:
1. 0.5A primary pickup, 5s time delay fed from CT in neutral. (Main 1)
2. 10V, 0.5s neutral overvoltage (NOV) fed from the neutral earthing trfr (essential a single phase distribution trfr). (Main 2)
Am now applying:
1. 100% stator EF in the SEL-700G using 3rd harmonic UV & NOV as well as retaining the 0.5A, 5s NEF relay (Main 1 protection).
2. 100% stator EF based on 20Hz injection (Main 2) as well as retaining the NOV function.
There has been quite some debate about what the trip times should be!
First thing is that any EF operation will result in Turbine Trip only. The GCB will then trip on RP in about 2s (soft landing) so one has to factor in around 2s additional delay to the trip time.
My view is a 2s trip for all EF functions which means fault will be removed in around 4s + CB time. Is there any point in a longer delay?
However, I am also keen to have a NOV element set to 80% of full displacement voltage which means it will detect faults at the machine terminals or 22kV network. Trip time now is around 0.5s. My concern here is for the voltage rise on the healthy phases with EF's. I wish to avoid a flashover with it's serious consequences.
Please also note that the two auxiliary trfrs as well as the GSU all have 22kV delta windings so the gen EF does not see EHV system faults.
I am keen to know what works for others.
Thanks in advance.