Bus coupler relaying 2

[Mbrooke]

What type of protection is required for a bus coupler (breaker 952) where both main bus bars are protected via a single (the same) differential zone? In such a case I'd assume its best to just leave the coupler out of the differential circuit altogether as there is no gain- but then that leaves the question about what to do with 10 CT circuits. In this case assume CTs are present on both sides of the coupler.

 

[davidbeach]

I've got no idea. That seems like a horrible layout from what I've seen, but I've never understood the point of double bus single breaker layout since there's no way of bypassing a breaker for maintenance. If both buses are in the same bus zone, I think you've just got a high priced switch with no interrupting function to perform. Using a bus differential relay the knows what position all the switches are in could provide some benefit at which point you could trip that bus coupler to avoid tripping the whole station for a bus fault.

I'd get rid of one of the two main buses and split the remaining bus into two or three sections with bus sectionalizing breakers between sections. But that's me.

 

[Mbrooke]

I've got no idea.

No worries, that makes at least two of us.

That seems like a horrible layout from what I've seen, but I've never understood the point of double bus single breaker layout since there's no way of bypassing a breaker for maintenance. If both buses are in the same bus zone, I think you've just got a high priced switch with no interrupting function to perform. Using a bus differential relay the knows what position all the switches are in could provide some benefit at which point you could trip that bus coupler to avoid tripping the whole station for a bus fault.

Agreed. You can add a 3rd transfer bus (as is the case here) which allows for breaker maintenance without taking a bay out of service.

The advantage to the second main bus is that the first one can be shut down for repair/maintenance without taking any lines out of service.

I'd get rid of one of the two main buses and split the remaining bus into two or three sections with bus sectionalizing breakers between sections. But that's me.

And alternating the lines outside the substation so generation / load and lines from differing right-of-ways end up sharing each bus section evenly.


In your experience with straight bus stations, how often does a bus section need to be taken out of service and for how long? In my observation its at least 1 in 10 year event.

 

[davidbeach]

When we wind up doing major bus work, like replacing cap-and-pin insulators with something newer, we'll take a bus outage, perhaps in the middle of the night, and cut the bus as necessary. The cuts will become the clearance points for the replacement work. Once that work is completed the whole bus is taken out again and welded back together, perhaps including a new cut for the next outage. In my nearly 11 years here we've not taken out the same bus more than once for bus work and given the relatively low number that have had to come out I'd figure it to more of a every 30-50 year event rather than a 10 year event.

Breaker-and-a-half is your best friend. Knowing of at least two total outages of double bus-double breaker stations due to multiple common-mode breaker failures (none of them ours) I don't buy the argument that DBDB provides more immunity to breaker failure than does BAAH. At least one of those examples would have only taken out a bay in BAAH and not the entire station; the other one may have had enough failures (ambient temperature related) that the configuration wouldn't have mattered. If the configuration didn't matter, the reduced breaker count for BAAH would have been money in the bank year after year for the many years since the station was constructed, but if BAAH could have kept one of the main buses energized the lower cost would have provided more value.

The diagram you posted puts 15 positions at risk for a single breaker failure. Ouch. Hopefully that is local breaker failure protection rather than remote backup protection. Trying to get that many remote positions to all clear in the right order seems to be asking more than I'd want to ask.

 

[RRaghunath]

Mbrooke, are you it is single differential zone?? Or is it single differential protection scheme?? In case of Low impedance differential protection scheme, a single relay / scheme can accommodate up to 4-zones.
If it is truly single zone, it is very strange. Considering that major expenditure is already incurred by way of large switchyard with three busbars, isolators etc. it would be appropriate even now to consider going in for at least two zone bus bar differential protection.
CT parameters etc. issues may not stand in the way in case of low impedance scheme.

Rompicherla Raghunath

 

[Mbrooke]

@raghun: Single differential zone. This comes up as a question often because of older stations being retrofitted with microprocessor relays but the existing system was only a single zone- (or even no zone in the case of remote clearing, though getting rarer). Smaller substations where the total loss of all bays is considered an acceptable risk also get this consideration.

 

[Mbrooke]

When we wind up doing major bus work, like replacing cap-and-pin insulators with something newer, we'll take a bus outage, perhaps in the middle of the night, and cut the bus as necessary. The cuts will become the clearance points for the replacement work. Once that work is completed the whole bus is taken out again and welded back together, perhaps including a new cut for the next outage. In my nearly 11 years here we've not taken out the same bus more than once for bus work and given the relatively low number that have had to come out I'd figure it to more of a every 30-50 year event rather than a 10 year event.

Id imagine all your remote substations are loop supplied (more than one sub), or at least tapping from two different bus sections? While outages from a bus fault are technically tolerable to some degree in my case, shutting down remote substations due to bus maintenance is avoided as much as possible.

30-50 years- consider yourself lucky.

One thing that takes considerable time in terms of bus outages are when a substation hits 50 years- needing replacement of the isolators. Though at that point you might be looking at a total re-build- yet not always the case come real world.

Breaker-and-a-half is your best friend. Knowing of at least two total outages of double bus-double breaker stations due to multiple common-mode breaker failures (none of them ours) I don't buy the argument that DBDB provides more immunity to breaker failure than does BAAH. At least one of those examples would have only taken out a bay in BAAH and not the entire station; the other one may have had enough failures (ambient temperature related) that the configuration wouldn't have mattered. If the configuration didn't matter, the reduced breaker count for BAAH would have been money in the bank year after year for the many years since the station was constructed, but if BAAH could have kept one of the main buses energized the lower cost would have provided more value.

It really is- but with a lot of existing substations EM relays force a variety of straight bus designs. Even newer less critical stations take cheaper configurations.


Personally where BAAH wins for me is that you can loose both bus bars and still keep your flow gate intact. This has been documented many of times to my knowledge- averting much greater outages.

The diagram you posted puts 15 positions at risk for a single breaker failure. Ouch. Hopefully that is local breaker failure protection rather than remote backup protection. Trying to get that many remote positions to all clear in the right order seems to be asking more than I'd want to ask.

This one has local BF- but here is the secret for those that do not: don't reclose on Zone 2 and Zone 3 at the remote ends unless the line comes live first. Works like a charm provided you do not have over-tripping of a line. Reverse zone 3 also works too. You would hit the floor if you only knew how many older substations across the globe (and even a few in the US) that do not have dedicated differential bus bar protection. Or even BF for that matter.

 

[Mbrooke]

And oh- Hey David, for fun of it take a look at what bus scheme is used for 1,200kv substations in India

https://www.slideserve.com/milly/1200-kv-ac-substation-basic-requirements-example-pgcil

Bet that would be BAAH no question asked in North America.

 

[bacon4life]

If the loads are normally on one of the main buses with the other main bus as the hot standby, Breaker 952 could have a settings group with a low set instantaneous element enabled to isolate the standby bus. If the loads are split between buses, perhaps a setting group for use when the bus differential is out of service?

I was warned by the older folks when I first started out that splitting the loads across two buses and relying on selector switches to correctly configure all lines into the appropriate bus differential zones would surely lead to a misoperation. Perhaps other have had better luck with dual differential schemes.

Bus maintenance seems at least as often as 1 in 10 years based on our double bus single breaker with breaker bypass substations. In addition to replacing the actual bus, we have had planned outage for PT replacement, structure painting, relay testing, and switch maintenance. A planned full clearance on some of our buses is not possible because of loops fed solely from a single source.

 

[Mbrooke]

If the loads are normally on one of the main buses with the other main bus as the hot standby, Breaker 952 could have a settings group with a low set instantaneous element enabled to isolate the standby bus. If the loads are split between buses, perhaps a setting group for use when the bus differential is out of service?

Do you mean a settings groups in the bus differential relays or a separate relay for the bus coupler like a 351?

I was warned by the older folks when I first started out that splitting the loads across two buses and relying on selector switches to correctly configure all lines into the appropriate bus differential zones would surely lead to a misoperation. Perhaps other have had better luck with dual differential schemes.

They are correct if your bus protection lacks a check zone. Check zone is your best friend. For EM (electromechanical) substations that takes place by having a separate set of relays which sum all the bays into a single zone, and then having another set of relays for zone 1 and yet another set of relays for zone 2 where CTs and trip circuits are selectively assigned to each set of zone relays. (Don't forget to also switch the BF trip circuits as well, if present) In order for a trip to take place, both a zone relay must call for trip AND the check zone relaying. That way if a misstep in switching causes differential current in either zone to call for a trip, the check zone will prevent any breakers from operating since there is no differential current in the check-zone. Even then things can still go wrong, ie an actual bus fault but the zones were left improperly configured clearing both busses.


All in all with EM relaying its much easier, simpler, cheaper and more reliable to just put both busses into a single zone.

If you really want to run a substation like that 50/50 your best bet by far are numerical microprocessor low impedance differential relays like 487Bs with position indicator switches on the bus isolators. Greatly simplified wiring and everything (zone switching, check-zone ect), is done internally to the relay including BF and electronic lockout if desired.


The 4th option is not having any dedicated bus bar protection- a practice that made single breaker double bus very attractive and very feasible in developing countries. Basically you set a step distance zone 1 that reaches through 10-20% of the shortest outgoing line left of the bus coupler with a 12 cycle delay, and a zone 2 (or another relay, so zone 1) that reaches 10-20% right of the bus coupler also with a 12 cycle delay. Line bays have a no delay zone 1, a 20-25 cycle delay zone 2 and a 65 cycle zone 4. Transformer bays also include delayed step distance elements that look through the bus bar or coordinated directional over-current elements. A fault on a bus causes the bus coupler to clear after 12 cycles effectively splitting the substation electrically in two, and then all the lines attached to the faulted bus clear either via a none relcosing revered zone 3 or remote zone 2 after 20-25 cycles- transformer bays in a similar manner. Times adjusted accordingly if breaker failure is used. Its arguable this method is even more stable then having any type of differential bus protection to begin with. Also this is good backup to when bus bar protection is implemented.

Of note- in this method (and its also been argued for the 3 prior methods, especially during a BF) I would recommend disabling the bus coupler during isolater switching in that if a bus fault occurred as an isolater is making or breaking one could have exceptionally high voltages and currents across the isolater contacts since the coupler is the first to open.


In your case, did these SBDB substations always come with bus bar protection or did they lack it at some point?

Bus maintenance seems at least as often as 1 in 10 years based on our double bus single breaker with breaker bypass substations. In addition to replacing the actual bus, we have had planned outage for PT replacement, structure painting, relay testing, and switch maintenance. A planned full clearance on some of our buses is not possible because of loops fed solely from a single source.

No full clearance- this is on your SBDB stations?

 

[bacon4life]

The bus tie breakers in our SBDB/WBP stations typically have 421s that can be used for line protection during bypass and for hot standby protection during normal operations. It sounded like your bus diff scheme was older, so I had assumed you would need a separate relay.

The SBDB stations originally had a high impedance differential scheme, which was disabled during switching, bypass, use of bus B or in one case, possibly never actually commissioned. The have now been replaced with 487Bs.

Yes, at some of our SBDB stations we have "radial" loops with embedded generation fed from the station. The terminals are such that breaking it into two smaller buses does not adequately balance load and generation under several scenarios. At other locations, we have been able split buses as David suggests. The major tradeoff is a required planned simultaneous outage of 5 terminals every 7-10 years for switch maintenance instead of the risk of an unplanned outage of all terminals. The second tradeoff is no longer having the ability to bypass circuit breakers. This concern is declining a bit since the maintenance of SF6 breakers is so much less than oil breakers. As part of the move splitting buses, we changed spare equipment strategies. We now have a trailer mounted breaker that can be installed in less than one day and we stock spare circuit breakers that can be installed within a couple of days.

 

[Mbrooke]

Older- but there is retrofit.


I'm having trouble picturing your setup. With the High Z relay, and 421 on the coupler, what did you do with the bus coupler CTs? A fault in standby would clear the other bus unless the coupler CTs were included somehow. Is this why you disable the bus protection?

 

[bacon4life]

In the normal configuration I think the High Z bus diff included the bus coupler CT so that the differential included only bus A. The 421 provide protection to the standby Bus B. When we switched to either bypass mode or having all loads on bus B, the differential was disabled.

In the original design, the bus tie relays needed to be adjusted for each different terminal it might backup. As we standardized on SEL400 series relays we now connect the bus tie CTs to the 2nd input on the relay. This lets us keep the line/transformer differentials active during bypass. The bus tie 421s then just have a generic settings group with backup protection

 

[Mbrooke]

In the normal configuration I think the High Z bus diff included the bus coupler CT so that the differential included only bus A. The 421 provide protection to the standby Bus B. When we switched to either bypass mode or having all loads on bus B, the differential was disabled.

That would make sense- allows for a fault on bus B to be cleared via the 421, but when differential current is on B, disabling the relay prevents a false trip. Personally I would use a 487B with a selection switch that simple add/removes the coupler CTs or creates two zones.

In the original design, the bus tie relays needed to be adjusted for each different terminal it might backup. As we standardized on SEL400 series relays we now connect the bus tie CTs to the 2nd input on the relay. This lets us keep the line/transformer differentials active during bypass. The bus tie 421s then just have a generic settings group with backup protection

Fill me in- how do you set the logic on the 400s regarding the CTs? I ask because before closing the bypass, the coupler and selected bay breaker are in series- so current is technically doubled when summed in the relay instead of the correct value when paralleled as in main and transfer.

 

[bacon4life]

The high Z relays have been replaced with 487Bs with check zone logic.

I don't remember the exact details of switching from the line terminal CTs to the bus tie CTs. The basic idea is only the X winding and the output to trip the line PCB is included settings group 1. The Y winding and the trip to the bus coupler is in settings group 2. One the bus selector switches transfers the terminal to the bus coupler and prior to closing the bypass switch, the relay settings group is changed.

 

[Mbrooke]

Makes sense.

Your isolators have position switches by chance?

 

[bacon4life]

Nope.

 

[Mbrooke]

Less that can fail