Stuck breaker satistics

[Mbrooke]

Would anyone happen to have statistics (or real world experience) regarding failure rates in 115 and 345kv SF6 dead/live tank breakers vs older bulk oil breakers? My understanding is that newer breakers are designed such with significantly more reliability and require far less maintenance when compared to "traditional" versions but unsure to what degree. I ask because stuck breaker and maintenance reduction is playing a role in new substation design.

 

[bacon4life]

The reduced planned maintenance windows for SF6 breakers has certainly influenced our recent substations designs for 115 kV substations. We have small batches from many different manufacturers, so unplanned maintenance (failure) rates seem to depend as much on the quality of the manufacturer as the insulating medium.

 

[DTR2011]

From the real world experience side, and assuming "stuck breaker" refers to a mechanical issue rather than what some would call in general "breaker failure". Assuming the mechanical side, I can offer the following observations.

Stored energy type varies and each system (air, oil or springs), has it's own unique operating conditions. I personally believe a modern spring charged mechanism is likely to perform best over time.

Old breakers may not have two trip coils, coil monitoring, modern relaying and breaker failure relaying schemes. Modern breakers do.

I am currently commissioning a Spring changed, 345kV, 63kA, (3x) single pole SF6 dead tank breaker. It has a number of built in alarm functions for many electrical and mechanical faults. It has 2 protection schemes (pole disagreement and 2 pole) for a "stuck breaker", which trigger alarms and the breaker failure relaying. I would suspect that the probability of a true catastrophic failure of this breaker is low over its lifetime, give maintenance responds to alarms.

On the other hand, I run into plenty of breakers that are 40+ years old, with Em relays that have lasted as long as they have without problem.

 

[davidbeach]

On the other hand, I run into plenty of breakers that are 40+ years old, with Em relays that have lasted as long as they have without problem.

And they will continue to do so right up to the instant that they don't any more.

Catastrophic failure of an oil breaker is ugly. I've seen a couple of those, after the fact, and in both cases I'm very glad to have been far, far away at the time of failure. Not so much an issue of the EM relays, but when the breaker has n fault interruptions in it, the (n+1)th interruption can be interesting. Now, if only we could know the value of n and the breakers fault interruption count in advance.

My gut reaction to the first of these was "and we let people in the substation....?" I never want to see one of these until well after everything is cold and the fire department long gone.

For the OP, life changes. We are installing better breakers and simultaneously providing far better protection against the day that the breaker doesn't work right. Oil breakers were almost always installed in single breaker applications and the new designs are much more likely to be dual breaker (ring or 1.5 breaker) installations. I think this is driven more by an increased focus on reliability than any concern about the reliability of the breakers being presently installed. Besides, to date we've seen many more breaker failures than our counterparts had in 1960 or 1970.

Part of the reason for significantly better protection against breaker failure is that it's free today. Practically every primary relay we install today can do breaker failure protection. Back in the day breaker failure protection required at least one additional relay and a bunch more wiring. That additional relay might require an additional panel. Besides, zone 3 would take care of everything, wouldn't it?

Back then I think breaker failure must have been considered an "if" event; today we know it to be a "when" event. As I sometimes have to remind people, it is Murphy that puts the food on my table and money in my bank account; if nothing failed I'd be unemployed.

 

[Mbrooke]

@bacon4life, How have designers changed the design of new substations? Ive seen the elimination of breaker bypass switches in new subs as one outcome.


@ DTR2011, Anything within the breaker itself that would prevent a trip on one or more poles excluding external relays and wiring. Breakers certainly appear less likely to fail alerting to certain aspects ahead of time.


@davidbeach, Makes sense, and great info. I do see redundant scheme taking over in new construction within North America and around the world, however I am curious if this the byproduct of growing load (a breaker failure being much more likely to produce unacceptable load flows or voltage collapse) or just a greater concern for reliability in general? I know of many older systems where ring bus was used for 345 kv and single breaker for 115/138kv with both being replaced by breaker and a half.


My primary concern is what are the odds of the breaker failing in its 40-50 year in service with proper maintenance, and how often is maintenance needed. Lower probability for failure makes keeping older single breaker bus bars in service more attractive, while eliminating breaker bypass in newer designs.

I did find this which to me was VERY eye opening. Designers have gone as far as eliminating disconnect switches and creating "discontenting circuit breakers" to serve both functions in order to save space and labor. Anyone who has had heard of this, is it for real of hyped marketing for special applications?

https://library.e.abb.com/public/b1... with minimum footprint 2013-03 Edition 2.pdf

 

[davidbeach]

Might work in locations where work practices don't require visible opens. They'd be a very hard sell in the North American market.

 

[Mbrooke]

In general dead tank breakers tend to be preferred over live tank, but these really throw a curve ball.

 

[bacon4life]

Our older substations were physically designed as double bus-single break with breaker bypass switches. The bus tie was used for both bypassing a breaker for planned maintenance and as a permanent spare since OCBs took quite a while to replace. Reconfiguring the bus tie and bus differential protection was a very big pain, so the stations were typically operated as a main & transfer scheme.

With the reduced maintenance of SF6 breakers, we now just open a bay for short duration of the maintenance. We have both a spare SF6 breaker and a mobile SF6 breaker that are available for emergency replacements with a few 10s of hours. With the higher circuit availability, we dedicated each terminal to a specific bus and installed much simpler sets of bus differential relaying.

In our latest ring bus station, we made the design choice to allow both generation sources to be adjacent instead of alternating gen/load/gen/load. Having breaker failure while one PCB was already open seemed a lower risk than having lines cross each other outside the station reach the alternate location.

Although ring bus/breaker and a half schemes are often touted as allowing maintenance of any component while maintaining service to all terminals, it really just allows for the CB mechanism maintenance. We have had a number of inadvertent trips while testing CTs and/or relays connected to one in service PBC and one out of service PCB.

 

[davidbeach]

Talk to me about that mobile breaker. Sounds like a fantastic idea, but I can't begin to picture how it would ever actually fit in. At this point I'm still trying to make a mobile set of VTs happen.

 

[Mbrooke]

"Our older substations were physically designed as double bus-single break with breaker bypass switches. The bus tie was used for both bypassing a breaker for planned maintenance and as a permanent spare since OCBs took quite a while to replace. Reconfiguring the bus tie and bus differential protection was a very big pain, so the stations were typically operated as a main & transfer scheme."


Sounds like each breaker had access to both buses via two disconnects? Did one relay or two relays give bus differential protection and what kind was used? Modern relays like SEL can easily be configured to accept dynamic current differential provided the disconnects have position switches where the relay has automatic awareness of the each disconnect's position knowing whether to add or subtract CT inputs from each dynamic buss zone. Although that is not to say I am not in the same boat as you lol. Most older substations have their existing relays with no position switches so substations designed as single breaker, double buss are often deliberately run as main and transfer with relays configured accounting for only one bus zone. As such a stuck breaker is a far greater concern since all circuits have to be cleared. Of course it is often justified with the reason being that a single relay summing both zones, or a typical application with older relays, is more likely to nuisance trip clearing both buses from service switching error (which is more likely to happen on a dynamic system) or the failure of a single relay will trip both buses as is, like this for example:

http://www.nerc.com/pa/rrm/ea/Lesso...202_Bus_Differential_Power_Supply_Failure.pdf

As for the buss coupler (or more precisely transfer bus breaker) what made re-configuring difficult? My practice is to omit differential bus protection altogether for the transfer buss and have the transfer breaker protect it. When not giving protection in place of another breaker I simply set a very low instantaneous current pickup (50P, 50G) to indicate that the transfer buss has faulted.



"With the reduced maintenance of SF6 breakers, we now just open a bay for short duration of the maintenance. We have both a spare SF6 breaker and a mobile SF6 breaker that are available for emergency replacements with a few 10s of hours. With the higher circuit availability, we dedicated each terminal to a specific bus and installed much simpler sets of bus differential relaying."

I thought about doing that many times (dedicating each circuit with two separate independent buss relays), but in your case what is the scenario for a bus fault? Do you keep half the circuits out until the bus is repaired or can you transfer the now de-energized circuits into the remaining buss while repairs are being made? With one buss zone while the entire station clears, all of the buss one disconnects can be opened and then all the buss two disconnects closed (often via SCADA) bring all circuit online rapidly while bus one is repaired. After repair, to transfer back all bus one disconnects are closed and then after, all bus two disconnects are opened. This is easily done without worry since only one buss zone is involved.


"In our latest ring bus station, we made the design choice to allow both generation sources to be adjacent instead of alternating gen/load/gen/load. Having breaker failure while one PCB was already open seemed a lower risk than having lines cross each other outside the station reach the alternate location."


PCB= power circuit breaker? Eliminating line crossing within a substation or other lines is a good goal both for safety and reliability. I tend to cringe when I see lines going over bays or multiple circuits to get to the other side. Ive seen cases where serve ice storms caused insulators to break with one circuit taking out two others.



"Although ring bus/breaker and a half schemes are often touted as allowing maintenance of any component while maintaining service to all terminals, it really just allows for the CB mechanism maintenance. We have had a number of inadvertent trips while testing CTs and/or relays connected to one in service PBC and one out of service PCB."


PBC? Sorry, Im not good with acronyms lol. Is it the middle breaker that trips open? Im just having a hard time grasping this. Then again breaker and a half is a design a least deal with it.

 

[davidbeach]

PCB = Power Circuit Breaker.

I think a lot of the design philosophy change is driven by the cost shift. In days of yore materials we expensive and labor was low cost. Now we'll happily pay more for materials if that can save us labor over the life of the installation. Dual breaker installation are easier to test, therefore have lower labor costs.

 

[Mbrooke]

But dont single breaker installations have fewer relays ultimately being easier to test?

 

[bacon4life]

yes, PCB = Power Circuit Breaker.
We originally purchased the mobile breaker as part of a 115/15kV mobile substations. The breaker sits on a trailer that has a set of 115 kV air break switches along with a dead tank breaker. For use with the mobile transformer, the umbilical cord gets plugged into relays on the transformer trailer. For use as just a spare breaker, we have a terminal board box where we reterminate all the circuits going to the broken circuit breaker.

RE Bus protection
-The key to dynamic bus differential is having position indication on bus selector switches. Retrofitting that on an existing 115 station was not cost effective. Also, retrofit position indication seem less reliable than circuit breaker position, and we have had issues with 52a/b contacts not matching breaker position.

We leave the bus section out of service until repairs can be made. On the overall system basis, we can tolerate any two "half" buses out of service with manual generation redispatch between outages. Using a single bus differential zone means having a "full" station bus outage with the chance for redispatch. Even worse is a full bus outage with other preexisting outages on the system.

It was debated whether to leave the disconnects in place to transfer lines. Ultimately we decided to remove the disconnects for two reasons. First, the disconnects would need periodic maintenance, which would necessitate a "half" bus outage and plus removing hot jumpers to the other "half bus". Second, the differential protection for the remaining bus would either be turned off, or have to be significantly more complicated.

Breaker & a half testing-
1. In locations where we use SEL 311L line differential relays, the CTs have to be paralleled outside the 311L. Even when one breaker is out of service, crews cannot short the CTS in the out of service breaker unless the 311L is taken out of service.
2. Prior to initial energization, we verify using primary current inject that the relay sees the correct currents. Even at locations with SEL 4XX relays that have 2 CT inputs, this is still a problem. If we can't take both lines out, one option is to open the trips on protection A and the CT inputs on protection B. Then reverse and repeat. While doable, it does add time, complexity and risk to the testing procedures.

 

[Mbrooke]

Thanks!

By bus selector switches you mean position indicating switches on the air disconnects?

Also, if you take both breakers out of service (one buss and the mid breaker) in breaker and a half does the risk/complexity still exist?

 

[bacon4life]

Yes, position indication of the bus air break switches. I attached a photo showings the switches to select bus A or bus B, tied by a few feed of horizontal pipe. To the very right is an oil circuit breaker with isolation switches and a bypass switch.

Testing CTs in the middle breaker affects both terminals. Testing CTs on the bus breaker affects the bus differential and the adjacent terminal. I suppose the bus might end up being a little easier to remove from service

http://files.engineering.com/getfil...-bc16-29b2a74159aa&file=10-07_disconnects.jpg

-David, the second link is a picture of our mobile breaker bypassing an oil circuit breaker.

 

[Mbrooke]

Thanks! Yes, I have substations exactly like that, although in some the bypass switch is tied to a transfer bus instead. I must ask, how come your substation does not have motor operators on the disconnect switches connecting to the bus? Also the bypass switches do not have arc whiskers because they do high current low voltage (loop) switching?

 

[bacon4life]

I thought manual operated switches was pretty typical at 115 kV. Even to have a mix of manual and motor operation at 230 kV. For the most part, the only motor operated 115 kV switches we use are part of automatic sectionalizing schemes, or on long taps to rural substations. At 230 kV, it seems to be based on whether an operator can physically operate the switch in a reasonably fast motion. In general, the substation switches are only operated as part of maintenance work where a crew needs a clearance. In this case, using a motor operator actually takes slightly longer as it has to be mechanically disabled as well as locked/tagged open. Any idea how much adding a motor operator in a new substation costs? The only numbers I have are for individual switches in transmission lines where the communications/SCADA portion was much more expensive than the actual motor.

The bypass switches just have arcing horns consisting of a the little ball on a 15 cm (6 in) rod. I don't recall every hearing of significant arcing while using a bypass switch. On our transmission line switches, we do use a variety of whips & vacuum bottles to deal with arcing from dropping energized line and splitting loops.

 

[Mbrooke]

Generally the status quo, but it depends on the utility. MODs are common for several reasons:

1. Enhanced worker safety (theoretically) since isolation switching is done by remote in the relay control building instead of at the switch itself.

2. Remote control SCADA ability to rapidly isolate a failed breaker and restore the bus its fed from.

3. In single breaker double bus stations, should a buss fault take place the MODs on the faulted bus are opened and all circuits are transferred into the healthy (un-faulted) bus re-energizing those circuits (all circuits in single zone bus differential stations) while work is done on the faulted bus. Once the faulted buss is repaired, circuits are transferred back.

4. In ring bus the the incoming line MOD is opened and the breakers are closed after the last line auto-reclose attempt fails. The reason for this is that should a second line outage take place while the ring is open, it can divide the reaming circuits from each other. This is a big concern where generation resources are on one side of the ring with generation limited load on the other.

5. Taking a breaker out of service while placing its associated circuit on the transfer buss can be done quickly with only one person rather then having several people in the yard to open and close disconnects in the correct sequence. Improper sequence will destroy a disconnect if opened carrying full load at full voltage.


6. Small stations without breakers which sectionalize transmission lines during faults.


2, 3 and 4 are to help meet reliability goals. While the system design consideration is such that the failure of any element including a buss or circuit breaker generally does not result in system instability and the loss of load over 80MW, it is highly advised that within 30-60 minutes the system be prepared to handle another contingency without encountering adverse operation there after. Within that time frame three options can prepare for a second contingency:

1. The system is of enough redundancy or capacity that no action needs to be taken as the second contingency will not result in negative operation afterward. Single transmission line faults or small generators as a first contingency are often such an example.

2. The faulted element(s) are removed or bypassed from the system within that time frame bringing the system to pre fault status.

3. generation re-dispatch, reactive resources (SVC, STATCOM, synchronous condenser), or similar are used to mitigate the negative impact of a second contingency. Even if stability wise the system can handle more then 1 contingency, lost load from any one or more contingencies must be restored within a certain amount of time.


Because a buss fault can be the literal equivalent of loosing 6, 8, 12 or more transmission lines; a failure of another major component like an auto transformer, large generator or especially another bus bar can trigger voltage collapse. Therefore its essential load flow through these transmission lines be restored within a short period of time (under 1 hour) which is accomplished via SCADA operation of MODs transferring everything to the remaining un-faulted bus. Of note, I know what your thinking, however the bus differential lockout is re-set via remote SCADA.


As for the MODs, becuase these do high current low voltage loop switching within the station (like placing a circuit on the transfer buss) or flopping to another buss while live, my understanding is that arcing whiskers can be damage as MODs are usually used in high voltage low current applications like transformer magnetizing current or buss charging current.





As for the price, my apologies, I don't remember off the top of my head, but when I get numbers you have my promise.