Setting Line Relays Near Generation

[davidbeach]

I'm wondering what others do when setting line relays at/near generation, looking onto the system away from the generation. In particular, setting backup protection - the traditional zone 3 or any overcurrent elements.

Historically we've essentially ignored the variability and set the relays like they were anywhere else in the system, everything in service and the generator impedance option set to subtransient. Well, it's easy enough to see that's probably right for a backup zone that will trip after 1 second. In the case I'm looking at at the moment, we have a 115kV line that has around 200MVA (nameplate) of generation behind it across 16 units at four plants. We might get most of that during the spring of a good water year, but by fall of a dry year it may be less than 70MVA of nameplate and only 4 or 5 units. Wildly different source strengths.

That 115kV line then terminates at a large station, on the low-side of a large (more than 300MVA) 30/115kV transformer. The standard criteria says that the third forward zone should be set to see past all that in-feed and reach the end of the longest line out of that station. That can wind up being further than the relay can reach.

Sequential tripping will get enough of the in-feed removed that the relay in question will reach as far as it needs to, but that can get rather ugly.

How do other people handle this type of situation?

 

[bacon4life]

The NERC hubbub about zone 3 relaying seems to say the risk of an overly sensitive zone 3 setting contributing to a widespread blackout is worse that the risk of relying on sequential tripping for backup protection.

We have some similar plants fed by a radial set of parallel 115 kV lines where we have to rely on sequential clearing. In the distant past, there was also a good likelihood of our generators tripping on low voltage prior to sequential clearing taking place. As we have improved the relaying, exciter and low voltage ride though capability, it is much less likely the generator would (incorrectly?) trip for a remote fault.

 

[davidbeach]

Is it worse to trip the generators on low voltage prior to the sequential tripping or on load rejection following the completion of the sequential tripping? ;-) Not sure the end result is much different. Oh, for a system with full redundancy, including batteries, and comms everywhere so we can get away from the needed for remote backup clearing.

NERC went way overboard on the zone 3 bit a decade ago. A whole bunch of protection holes were created that took time to fill back in again; faults that would need to burn back toward the source before anything would trip. Bah humbug. Load encroachment was/is a much better solution to that mess than no remote backup.

 

[krisys]

Zone-3 setting would always have enough time delay, say typically, 500 ms. By that time the sub transient would die down. Since it is a near generation plat, probably transient would be more appropriate.

As there is enough time delay for Zone-3, the risk of cascade tripping is a lower probability.

 

[marks1080]

We don't typically do anything special. Set the line as any other line. If you're using DCB than set the zone 3 approrpiatly. The generator will probably have a main output transformer that will make setting the zone 3 (looking back into the plant) easier. This obviously gets more difficult with traditional distance schemes in conjested areas with short lines/3-ended lines, etc... but again, we treat those situations no different if its at a generating station or not. And there are always one off scnearios where ya just can't get the POTT/PUTT/DCB schemes coordinated perfectly, so you chose which you want to sacrafice between reliability and dependability.

We've been using 87 more and more on HV lines and it makes things a lot easier.

 

[krisys]

Alternatively, use the fiber optic based feeder differential protection scheme for the additional system security. Specifically, if the line length is too short. Now a days OPGW (Optical Ground Wire) conductor is widely used for the Over Head Line (OHL) shielding. So, the Fiber Optic (FO) link is not a problem.

 

[davidbeach]

Oh, but were it that easy. I'm not particularly worried about the protection of the line itself; the concern is with the longer reaching, slower elements used to back up the protection at the remote terminals. Line diff is a great solution for the problem I don't have (will be able to turn on the diff in a year or two anyway) and does nothing for remote back up. I've seen enough go wrong enough different ways that I don't believe that a series of lines with line differential provides an adequate level of overall system protection.