Why do some utilities use a totally stand alone relay for breaker failure and reclosing? (Point on wave switching not required) What advantage if any is there? I've always thought it best to just let one relay (with a single back-up of course) do everything- makes more sense.
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!
Breaker Failure via Separate Relay 9
- Thread starter Mbrooke
- Start date
-
1
- #2
Reclosing relays and BF relays were separate devices in the days of electro-mechanical relays out of necessity. I don't see a reason to have separate reclosing relays if a digital protective relay with reclosing capability is used. It still makes sense for BF function to be separate from a relay that might have failed to operate, IMO, although it can certainly be built into a digital relay. This makes more sense if there are redundant relays, trip coils, batteries, etc.
- Thread starter
- #3
By relay that failed to operate you mean both primary and secondary distance relays (421,311C) that are capable of initiating breaker failure (stand alone 351)? Thats interesting.
On a similar note is there any down side of having both primary and secondary relays with BF enabled inside each one?
On a similar note is there any down side of having both primary and secondary relays with BF enabled inside each one?
-
2
- #4
davidbeach
Electrical
In the world of full redundancy, including dual batteries and dual trip coils, external functions, such as breaker failure, would also have to be fully redundant. If each relay does its own breaker failure you have full redundancy. Also, if each relay does its own breaker failure and has no external BFI input the current detector can be set low. Any relay that has an external BFI input should have the current detector set well above load current so that a spurious BFI signal doesn't result in clearing the bus.
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
- #5
David, your POCOs is very lucky to have you! Not many are capable of your practical yet secure way of thinking. Separate BF relay, replicated only once, and current pickup set such that load current can trip it. Don't think I can bring about change, but I'd sure like the POTT/87 devices to also handle the BF.
If the 411L is also doing the BF, should it also be enabled in the backup 311C relay? I'm reading a lot saying you need BF only in one relay and that its typically not replicated.
If the 411L is also doing the BF, should it also be enabled in the backup 311C relay? I'm reading a lot saying you need BF only in one relay and that its typically not replicated.
davidbeach
Electrical
The only difference in programming between our ‘A’ and ‘B’ relays is in the event recording/triggering conditions. A few long events triggered by TRIP on the ‘A’ and a lot more, shorter, events triggered by multiple conditions on the ‘B’.
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
- #7
davidbeach
Electrical
Why would they fight?
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
- #9
davidbeach
Electrical
Breaker interrupting times are shorter than breaker closing times. The relay only initiates reclosing if the breaker was closed at the time the relay issued a trip. So there's a 3-cycle or so window where they can trip at different times and then attempt to reclose at different times. Neither relay is going to have a problem with that. If only one relay issues a trip then only that relay will attempt a reclose.
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
- #11
thermionic1
Electrical
Once a breaker is open, further trips don't do anything. The breaker internal 52a opens the trip circuit. Once a breaker is closed, further closes won't do anything. The anti-pump circuit in the breaker prevents it. These circuits are integral to every MV/HV breaker I've ever seen.
As an aside, consider human performance issues when having everything all in one box. I attended a conference held by our transco and as far as inadvertent trips go, accidentally sending a BFT to the 86B on a single bus configuration was very high for several years. This can happen because someone forgets to pull the line relay BFT and is doing SCADA checks or similar. The story I heard was the BFT cleared the bus which was also the feed to the level 1 trauma center. Of course, their emergency gens failed to start, so a complete blackout to a major hospital. That was years ago and the hospital now has separate HV voltage feeds, but these things can happen. I was overseeing a utility relay tech do similar SCADA tests and had to jump in and pull the FT switches when this guy almost made a similar mistake a few years ago.
If you have line protection separate from BF protection, if the breaker is open (no current to BF Relay), all the BFI's in the world won't cause an issue.
If the relay is in 79 Cyc, the breaker is already open.
As an aside, consider human performance issues when having everything all in one box. I attended a conference held by our transco and as far as inadvertent trips go, accidentally sending a BFT to the 86B on a single bus configuration was very high for several years. This can happen because someone forgets to pull the line relay BFT and is doing SCADA checks or similar. The story I heard was the BFT cleared the bus which was also the feed to the level 1 trauma center. Of course, their emergency gens failed to start, so a complete blackout to a major hospital. That was years ago and the hospital now has separate HV voltage feeds, but these things can happen. I was overseeing a utility relay tech do similar SCADA tests and had to jump in and pull the FT switches when this guy almost made a similar mistake a few years ago.
If you have line protection separate from BF protection, if the breaker is open (no current to BF Relay), all the BFI's in the world won't cause an issue.
If the relay is in 79 Cyc, the breaker is already open.
- Thread starter
- #13
Hadn't thought about it that way in regards to BFI not activating with a separate relay. Though thats where test switches come in... but yes, human error is inevitable.
Could you ever get two recloses from a scheme where both primary relays are set to one attempt?
Regarding the hospital I'm even more curious. Do you know why multiple gens failed to start?
Could you ever get two recloses from a scheme where both primary relays are set to one attempt?
Regarding the hospital I'm even more curious. Do you know why multiple gens failed to start?
-
1
- #14
davidbeach
Electrical
It does require identical reclosing logic in both relays to make it work. In other words, it requires a matched set of relays. Almost everywhere at transmission we have matched relays and I've never seen any extra operations there.
But at the distribution level we've had to make sure that anyplace with unlike relays only one should be expected to do anything if it is working. We've had various combinations over the years and any differences in reclose logic can result in the two relays sharing the duty and getting too many operations.
Subtle differences in reclose stalling and reclose supervision can result in one relay getting the breaker closed before the other is ready to try. Then one relay is at shot 1 while the other is still at shot 0. So the second trip is a second instantaneous by the relay at shot 0. We've seen 351/551 and 351/751A combinations fail. 351/551 was a designed condition; a 351/751A combination is the result of a breaker that originally had a DPU and a BE1-51 having the BE1 replaced with the 751A to address DER connections and then the DPU failing and being replaced with the 351 since a second 751A wouldn't fit in a DPU hole.
But that's distribution and we've solved it by having the "better" relay trip quicker and do all of the reclosing and let the "lesser" relay provide single shot backup protection. Now distribution is also a matched pair (has been for years now) and both can function autonomously with no conflicts.
By the way a 421/311C combination is absolutely ridiculous and never had any justification. We did do a handful of 421/311L combinations and regret them all; the differential just wasn't worth it. Can't see any justification for two different distance relays at anytime in recent memory. Two different relays just means twice to opportunity to have an unnecessary trip. 30 years ago there might have been a concern about that new fangled relay actually tripping; those days are long gone. Well over 90% of all reported misoperations are overtrips; there just isn't a problem with failures to trip. Concentrate 100% of the design time on getting one relay type just right rather than putting 50% of a limited resource on one relay and the other 50% on a different relay. That 421/311C combo is probably less problematic than any SEL/GE combo (who has the resources to throw away on thoroughly understanding two different systems with very different corner cases?) but a horrible waste of resources. Even on a first cost basis, two 421s or two 311Cs would be less expensive because only one set of settings have to be developed and maintained. Two sets of settings completely consumes any perceived hardware savings. Then there's the added O&M costs to be absorbed over the life of the installation. Based on history, lowest first cost does not result in lowest lifetime cost. For a developer who's going to flip a property in a couple of years first cost is everything; but for a utility that's going to own it forever and replace it at some point the only thing that matters (or should matter anyway) is life cycle cost. We have some lowest first cost installations; hate them all and I'm sure we've long since spent all the savings in increased O&M costs. Saved a buck 25 years ago and we've spend 15-25 cents a year since because of it. Long since ceased to have made any sense.
I’ll see your silver lining and raise you two black clouds. - Protection Operations
But at the distribution level we've had to make sure that anyplace with unlike relays only one should be expected to do anything if it is working. We've had various combinations over the years and any differences in reclose logic can result in the two relays sharing the duty and getting too many operations.
Subtle differences in reclose stalling and reclose supervision can result in one relay getting the breaker closed before the other is ready to try. Then one relay is at shot 1 while the other is still at shot 0. So the second trip is a second instantaneous by the relay at shot 0. We've seen 351/551 and 351/751A combinations fail. 351/551 was a designed condition; a 351/751A combination is the result of a breaker that originally had a DPU and a BE1-51 having the BE1 replaced with the 751A to address DER connections and then the DPU failing and being replaced with the 351 since a second 751A wouldn't fit in a DPU hole.
But that's distribution and we've solved it by having the "better" relay trip quicker and do all of the reclosing and let the "lesser" relay provide single shot backup protection. Now distribution is also a matched pair (has been for years now) and both can function autonomously with no conflicts.
By the way a 421/311C combination is absolutely ridiculous and never had any justification. We did do a handful of 421/311L combinations and regret them all; the differential just wasn't worth it. Can't see any justification for two different distance relays at anytime in recent memory. Two different relays just means twice to opportunity to have an unnecessary trip. 30 years ago there might have been a concern about that new fangled relay actually tripping; those days are long gone. Well over 90% of all reported misoperations are overtrips; there just isn't a problem with failures to trip. Concentrate 100% of the design time on getting one relay type just right rather than putting 50% of a limited resource on one relay and the other 50% on a different relay. That 421/311C combo is probably less problematic than any SEL/GE combo (who has the resources to throw away on thoroughly understanding two different systems with very different corner cases?) but a horrible waste of resources. Even on a first cost basis, two 421s or two 311Cs would be less expensive because only one set of settings have to be developed and maintained. Two sets of settings completely consumes any perceived hardware savings. Then there's the added O&M costs to be absorbed over the life of the installation. Based on history, lowest first cost does not result in lowest lifetime cost. For a developer who's going to flip a property in a couple of years first cost is everything; but for a utility that's going to own it forever and replace it at some point the only thing that matters (or should matter anyway) is life cycle cost. We have some lowest first cost installations; hate them all and I'm sure we've long since spent all the savings in increased O&M costs. Saved a buck 25 years ago and we've spend 15-25 cents a year since because of it. Long since ceased to have made any sense.
I’ll see your silver lining and raise you two black clouds. - Protection Operations
- Thread starter
- #15
David, I'm surprised you would say this. Its cool, but do you remember Astoria? I would never do two 421s or two 311Cs. The back-up or "B" relay must a different model. Same with static busbar protection, there is always a 487 primary with a 587Z back up. 351/751 feeder relays, 487E/387 transformer protection. Relays must be different. 551C/501. ect.
Relays aren't matched, hence my concern.
Relays aren't matched, hence my concern.
Another case where the BF relay should not be in the distance relay, is where you have a ring bus, or breaker and a half scheme. Can get very confusing when four distance relays try to do breaker failure for one breaker.
By having different relays for the functions, the logic becomes more understandable to engineers and anyone trying to troubleshoot.
And also the testing thing.
Have you had problems with common mode relay failures? I haven't seen this with SEL relays. So why would you want different relays? Maybe you like doing the settings work twice.
And why would anyone need a second breaker failure relay? Is the breaker failure to be used only if the breaker actually fails? Maybe you need to buy better breakers, or maintain them better.
I think maybe people might be over complicating things, and causing more problems than they are solving.(Why not three or four protection schemes?)
By having different relays for the functions, the logic becomes more understandable to engineers and anyone trying to troubleshoot.
And also the testing thing.
Have you had problems with common mode relay failures? I haven't seen this with SEL relays. So why would you want different relays? Maybe you like doing the settings work twice.
And why would anyone need a second breaker failure relay? Is the breaker failure to be used only if the breaker actually fails? Maybe you need to buy better breakers, or maintain them better.
I think maybe people might be over complicating things, and causing more problems than they are solving.(Why not three or four protection schemes?)
- Thread starter
- #17
Aren't most utilities going for BF and 21 in one device even where using BAAH and Ring? I'm hearing the words confusing, but, I'm having trouble as to picturing what exactly is confusing tbh. Less wiring and fewer relays makes it easier as first glance.
Yes I've never seen SEL common mode failures, but what if it did happen by chance? A batch of relays that just happened to have a defective component by whom ever is supplying SEL?
Yes I've never seen SEL common mode failures, but what if it did happen by chance? A batch of relays that just happened to have a defective component by whom ever is supplying SEL?
davidbeach
Electrical
Ring and BAAH each relay does its own breaker failure for the breakers it trips. Easy and simple.
The lesson of Astoria seems to have been missed. It matters not how many different relays are used if none of them are set correctly. The failure of the remote end to detect the fault through the reactor is the primary failure. That two communication power supplies both reacted badly to the same DC transient should have had no more impact than slowing down the clearing by 15-25 cycles.
We're never seen both relays on a position fail at the same time, even when dealing with families of relays that are clearly at end of life. No, the greatest risk of common-mode failure occurs during the schematic design and setting development process. The more the Engineers can concentrate on a single design, a single relay with its idiosyncrasies, and a single setting package the less likely that position will have a misoperation. And if it does misoperate it's easier to determine if that misoperation is limited to that position or if it is systematic problem.
I’ll see your silver lining and raise you two black clouds. - Protection Operations
The lesson of Astoria seems to have been missed. It matters not how many different relays are used if none of them are set correctly. The failure of the remote end to detect the fault through the reactor is the primary failure. That two communication power supplies both reacted badly to the same DC transient should have had no more impact than slowing down the clearing by 15-25 cycles.
We're never seen both relays on a position fail at the same time, even when dealing with families of relays that are clearly at end of life. No, the greatest risk of common-mode failure occurs during the schematic design and setting development process. The more the Engineers can concentrate on a single design, a single relay with its idiosyncrasies, and a single setting package the less likely that position will have a misoperation. And if it does misoperate it's easier to determine if that misoperation is limited to that position or if it is systematic problem.
I’ll see your silver lining and raise you two black clouds. - Protection Operations
- Thread starter
- #19
Right, but my understanding is that the primary protection was not supposed to see past the reactor, hence 25-30 cycle clearing was not possible. Communications was essentially the primary protection. Both communications processors were the same, and the same problem effected both of them.
Relay engineers and protection groups should know how to set multiple relays without sacrificing anything.
Relay engineers and protection groups should know how to set multiple relays without sacrificing anything.
-
1
- #20
davidbeach
Electrical
We may simply have to agree to disagree, but Astoria was primarily a settings failure. It is unconscionable to implement settings that can't clear all (hard) faults in the zone. Comms is simply to speed things up.
I could "set" a GE line relay today, having never worked with one before. But those settings would be simply assuming that everything can be taken on face value. Pretty sure that's not actually true. There are many subtleties in relay operations, particularly around the edges. Pretty sure the GE relay doesn't use the SEL directional elements. It may, in fact, be impossible to have both always agree on directionality in the corner cases. Those corner cases are where the misoperations happen. Much better to put 100% of the effort into "continuing education" about one relay/relay family, than to dilute that effort among multiple relay types. Not just talking about the understanding of the the relay that someone turning out multiple settings per year, but also talking about the understanding necessary for someone to produce the setting criteria. There's less than a couple dozen reach settings, those are easy; it's how they interact with each other that takes much more effort.
I’ll see your silver lining and raise you two black clouds. - Protection Operations
I could "set" a GE line relay today, having never worked with one before. But those settings would be simply assuming that everything can be taken on face value. Pretty sure that's not actually true. There are many subtleties in relay operations, particularly around the edges. Pretty sure the GE relay doesn't use the SEL directional elements. It may, in fact, be impossible to have both always agree on directionality in the corner cases. Those corner cases are where the misoperations happen. Much better to put 100% of the effort into "continuing education" about one relay/relay family, than to dilute that effort among multiple relay types. Not just talking about the understanding of the the relay that someone turning out multiple settings per year, but also talking about the understanding necessary for someone to produce the setting criteria. There's less than a couple dozen reach settings, those are easy; it's how they interact with each other that takes much more effort.
I’ll see your silver lining and raise you two black clouds. - Protection Operations
- Status
Similar threads
- Question
- Question
- Question