Bus Differential & Lockout Relay 5

[Yo121]

I'm trying to get an answer as to why use a lockout relay (LOR) on a 487B low-impedance bus differential relay that uses dynamic zone protection. In my opinion, this isn't necessary unless the utility wants someone to be forced to go to the substation and inspect the bus before resetting the LOR. Then again, some LOR's are able to be reset electronically via SCADA, without anyone going to the station to see what may have caused the fault...this is particularly advantageous for remote substations such that service could possibly be restored faster if a re-close attempt is made. In my opinion, a utility operator could accidentally close breakers on a bus, transformer, etc., if he can do it in the control room via SCADA; hence, the use for the LOR to block close the breakers until it's physically, rather than electronically, reset. Also, according to GEC Measurements, Protective Relays Application Guide, General Electric Company p.l.c. of England, 1975, the vast majority of bus faults were recorded to be an insulation flashover, at 21%; breaker failure, at 15%; switchgear insulation failure, 17%; with disconnect open or grounded and safety ground leads still left on bus after maintenance, both at 11.5%; other insulation failure (maybe mechanical failure) amounted to a mere 7% of bus faults recorded....very, very rare with the fact in mind that bus faults are very rare in the first place. So, if the breakers are re-closed without inspecting the bus, additional, possibly serious, damage could be done to the bus; however, it's not likely serious damage will be done upon re-closing and also the relay should be programmed or designed to reset automatically before a re-close is made.

It seems to me that LOR's are used to lockout buses for the following reasons: High impedance diff relay's don't have enough outputs to trip each breaker individually; high impedance diff relays MOV's cannot sustain more than maybe 4-5 cycles of secondary fault current (clamping the voltage across the high-impedance element) before MOV failure (LOR contact shunts around MOV/high-impedance relay elements); finally, LOR for bus protection are used where there's a combined tranformer/bus diff relay, where the fault could be internal to the transformer and not the bus; secondary to all these above listed requirements, again in my opinion, is requiring a person to go and inspect the bus before attempting re-closing, but I can see the use of LOR's as a safety measure for personnel on site after a bus fault occurs so they aren't shocked again if someone re-closes right after the fault...this could be mitigated by programming SCADA to block close after a bus fault, though. Keep in mind a bus fault is a major event, and I doubt most operators would simply, haphazardly close all the breakers connected to a bus, even if they could electronically reset the LOR and attempt a SCADA close on the breakers terminated on the bus.

My personal opinion on the matter is it's not necessary to lockout a bus upon a bus fault, using a low-impedance differential relay....especially the SEL-487B relay with dynamic zone selection and high-speed output contacts (LOR operates slower). Then, if a utility desires, the SCADA controls can be programmed to lockout via feedback of breaker b status contacts (52b) and an output(s) from the 487B relay (or using the relay word bits on a comm protocol like DNP3). The only time I'd voluntarily lockout a bus--if the utility or industrial substation owners didn't care one way or the other--is for high-impedance diff relays, combined transformer/bus diff protection, and breaker failure (using a breaker failure LOR). What say you and why?

 

[davidbeach]

No. Lines can be reclosed, every other component of the electrical power system gets one chance; if it faults a lockout is tripped and somebody goes there to inspect, analyze, and evaluate. That's it; that's the whole point of an LOR; force that human evaluation. Electrically resetable LORs are an abomination; the fact that something can be done is no reason for it ever being done.

A bus fault is a bus fault, I don't see how a bus trip from an SEL-487B is any less of a bus trip from an SEL-587Z. Bus trips, bus gets locked out. Simple, clean, and generally minimizes the system impact to any load not served directly from the tripped bus. Just like there are a few bus trips that could be successfully dealt with by a single reclose there are line trips to lockout that might have held if we'd given it one more reclose, but where does it stop? Pick a number, let the protection system get to that point, and then send somebody there. Blind SCADA closing after the intended number of recloses (including zero) is just asking for trouble.

Are there bus faults that could be cleared with a single reclose attempt? Sure, no doubt about it, but what's the fault mechanism and what's the likelihood of success? My understanding is that FPL (at least used to) routinely recloses once on bus faults due to a long history of bus misoperations caused by close-in lightning strikes. We, on the other hand, have a transmission bus trip every few years. Distribution bus trips, unfortunately, are rather more common and generally result in enough damage that no one would ever envision anything less than a forced visit to the station to catalog the damage prior to repair and then restoration.

 

[Yo121]

I get it, fault once then done until inspection time. Sure, but many bus faults are temporary and they're very, very rare compared to other power system apparatus faults. Also, it seems to me that the SEL-487B response is slowed by using a LOR, especially if high speed outputs are used. Unless it's desired to break the close circuit of the circuit breaker, which can also be done using a multi-processor protective relay output, you don't need a LOR for a low-impedance bus diff relay. On the other hand, LOR's are ESSENTIAL for high-impedance diff relays; e.g., the SEL-587Z. You literally cannot install a 587Z relay without a LOR. This is because the energy the MOV's would have to dissipate is greater than what they can practically be designed to handle over the duration the secondary CT current flows for an internal bus fault.

Let me give an example: A breaker may operate in 4 cycles or more...generally a higher voltage stations, they're now at 2-3 cycles clearing time. Yet, traditionally, they were at 5 cycles or even 8 cycles clearing time. Also, keep in mind that there's practically at least a 1.5 cycle delay for a high impedance diff relay to trip, upon a bus fault...low impedance (487B) can be less than 1 cycle. Even with 3 cycle breakers, we're talking 4.5 cycles minimum, easily exceeding what the MOV's inside the relay can handle. So, the LOR is absolutely necessary for a high-impedance diff relay, because a contact is wired to shunt away current from the high-impedance relay element to protect it. This is very much not the case with a low-impedance relay, like the 487B.

Also, LOR's are used to lockout a machine or other piece of equipment to do maintenance/inspect it or to lock it out upon an internal fault and prevent an expensive piece of equip from getting further damaged. What maintenance or inspection is required for a bus, besides maybe cleaning insulators in high-pollution areas? I can see the argument that a bus should be locked out, by preference, because it's a node of the power system. Even so, just open the breakers and open their disconnect switches, problem solved...and that's probably safer than using a mere LOR that someone could accidentally reset by hand on site and close in a breaker they didn't intend....unless there's a physical lockout/tagout procedure implemented on the relay panel after a fault/during maintenance.

Finally, SEL promotes not using LOR's for bus trips on the 487B. They say you can trip the breaker directly with the 487B. They promote breaker failure lockout relays, only. So, the point isn't for a LOR to do a one and done, upon being tripped by a relay for a fault; the point I'm making is why is a LOR necessary for low-impedance bus fault protection? I don't think it is, except to lock out the bus upon a fault to prevent reclosing breakers (if that's company policy). No, what I'm looking for is someone to tell me something technical, like my example of the high-impedance MOV's, to justify its use in a low-impedance bus diff 487B application. I believe it's a utility/plant owner policy issue due to their previous experiences, not a protection issue concerning the 487B relay.

 

[cranky108]

An LOR will operate in a quarter of a cycle, so it only slows things down by that much. How much do you gain by the increase in speed for this fault clearing?
Do you also use cycle and a half breakers?

How many bus faults do you have in a year? Is it a problem to inspect the vast number you have each year?
I admit I work with a small system, but in the past 5 years we have had 1 actual bus fault, and it would not have held on a reclose. Besides if you do your bus shielding right lightning should be a minimum issue.

Just to be clear, our company policy is to fire anyone who resets a lockout by accident, and without the permission of the system operator. Everyone knows that.

Why did you use the 487B over the 587Z? It was not because it cost less. Did you actually need the zones? Or was it the Gee Wiz factor of your ego?

After a bus differential operation we require each breaker disconnect be opened, and the bus inspected, including Arresters, PT's, etc.

The lockout is a visual indication that you can't see as well on a relay, as well as requiring someone to go to the substation. We don't use resetable lockouts as they are slower, and cost more.

In most cases, what you would spend extra on the 487B relay, and labor for the applying it is not worth it. The exceptions are for EHV where you have cycle and a half breakers, and where you need the different zones.

 

[davidbeach]

We're using dual 487Bs on transmission buses for PRC-005 considerations, but find the 587Z completely suitable for distribution buses.

 

[cranky108]

I suppose because you are using the 487B also as breaker failure. Sort of a two in one.

We use the 351 for breaker failure and sync-check. As a two in one.

A little different way of thinking.

 

[Yo121]

I'm using the 487B on a single breaker, double bus substation. This will use six zones of protection, one per phase for each bus. There's significant local generation present that can be switched onto either bus. So, dynamic zone switching is necessary, with some sort of check zone or a directional overcurrent supervision. I'm using high-speed contacts to reduce stability problems with delayed tripping of such a major bus, enabling the relay to trip a breaker in less than 1 cycle after initiation of a fault...the 487B is highly sensitive to a fault (internal and sometimes external), and the challenge is really to de-sensitize it enough so it doesn't nuisance trip while maintaining enough sensitivity. In addition, the 487B relay is being used as a sync check relay via mirrored bits from a 451 relay. That is, A contacts are in the close circuit from the 487B relay, because the 451 relay outputs are all in use. The A contact from the 487B relay knows when to close on synch check via mirrored bits from the 451. Finally, the 487B relay is being used as a station breaker failure relay with external lockout relays for breaker failure.

The decision was made to go with the 487B relay, because it is more reliable and easier to implement than some complex control scheme, with 587Z's, to remove/add CT's from it with auxiliary relays as the zones of protections change. So, the cost was well worth it, and the savings from not using bus lockout relays and additional breaker failure relays, along with a complex auxiliary control scheme that could cause many headaches, were realized. After careful deliberation, we've decided to use some 487B contacts as in a lockout relay. That is, we will hold closed a 487B A contact in the close circuit, continually, and the 487B will directly trip the breaker. When the 487B trip outputs assert, the A contact in the close circuit will open and block close the breaker. The A contact also serves as a "fail-safe" contact to block close the circuit breakers, if the relay fails for some reason. Finally, the 487B relay has a big button on the front panel that says "TARGET RESET" that will need to be pushed to unlatch the trip output and latch the block close outputs, similar to resetting a lockout relay. Thanks for everyone's input so far.

 

[cranky108]

Is this relay a single point of failure issue? Is there a logic complexity issue? What about testing?

There is something about keeping it simple, that sort of sounds like you have too much in the 451 relay, and have to use the 487B to provide the outputs.

At times the complexity is required, but most times it is not needed. The lockout in the blue box is not an idea many of us want to work with, and is too far from simple to justify.

 

[Yo121]

It was use the blue boxes or slew after slew of auxiliary relays that have a reputation for failure and complexity in this type of control scheme. We are using primary 487B, with a backup 487B. Yes, they're both 487B relays and not different manufacturers/models like GE B90 for a backup, but the utility was okay with that and SEL only makes the 487B that's acceptable for this application...there's also generally no requirement that I'm aware of for the primary/secondary relays to be of different manufacturer or even different models. They're on separate batteries, trip coils, etc. However, trust this when I say that using the 487B is simpler than using a 587Z. Using the 487B as breaker failure is also very simple and easily testable. Using the 487B mirrored bits comm with the 451 is also very simple and easily testable. Programming the 487B as a lockout relay, central to one relay, remember there's a backup 487B doing the same thing, is also very testable and simple...and using a form a contact to block close the breaker (latching the form a contact close otherwise) is a type of fail safe contact that doesn't allow the breaker to close if either 487B relay is out of service...which is desirable for this application. It sounds crazy and complex, but combining all this into one relay, and duplicating it with another backup 487B, is actually simpler and less costly and not as complex as it may seem. All that has to be done is document what the relay is doing, properly.

All I'm doing, really, is removing hardware and replacing it with software....which quite frankly is the future/present state of substation protective relaying that's actively being implemented, to various extents, around the world. These numerical protective devices are extremely reliable, often times more reliable than solid-state or electromechanical relays. Also, it's acceptable to lockout out a bus with a microprocessor protective relay contact, as in an SEL relay, according to C37.234 Sec. 8.6.1...of course as long as the utility/plant owner doesn't object. I also prefer using software when I can get away with it than hardware...with redundancy of course. Software has really reached a point where it's either just as reliable as hardware or even more reliable than hardware. Software also gives huge scheme flexibility and lowers install costs. I liken it to using a PLC vs. using cabinets full of relays in process control applications (like a food & beverage production plant). Now, software needs cyber security, of course, but that's implemented, too. In my opinion, from my experience, it's really a matter of time before everything in a substation is implemented in several "black boxes", programmed with software, and even electroswitch lockout relays are no longer used. Touchscreen HMI's then replace the visual experience of switches rolled, etc. More and more utilities are going/have gone this route.

 

[breadboard]

I've seen plenty of extremely reliable microprocessor relays fail, old fashioned LORs do have the advatage of not needing to be programmed at all.

 

[rcw retired EE]

Presently we are working on 8 breaker, breaker and a third 230kV yard, and a 5 breaker, breaker and a half 500kV substation for a large east coast utiltiy. Their standard is to automatically reset lockout relays 2 seconds after the fault current is zero and the 87T trip signal is off. Same logic is applied to line, bus and "Stuck Breaker" schemes. Lockouts can also be reset via SCADA, a local HMI or manually. The lockouts only inhibit closing, they do not trip the breakers. The 487B and B90 high speed output contacts trip the breakers independently. Both primary GE B90 and back up SEL487B have their own lockout.

The auto reset feature might make sense on failed breaker logic. When a Stuck Breaker (breaker failure) is detected, the F60 or SEL451 SB relays trip teh local brekers, send DTT to remote ends and rolls the lockout to inhibit closing adjacent breakers. When current = 0A, both SB relays signal the breaker disconnects to open on either side of the failed breaker after a short time delay. The lockouts are then automatically reset by SB relays' logics. I assume this is to allow fast automatic restoration after a breaker failure.

Personally, I like the lockout as a LOCKOUT that should not be reset except by someone with eyes on the protected or damaged equipment. Cusotmer is always right.

 

[davidbeach]

Yep, the "Customer is always right", but they can sure provide some good stories to tell around the bar at relay conferences. Particularly after they have the messy event that proves the old way was better than doing something new and different just because it's possible. I've pushed lots and lots of change here, but there are somethings, like the good old LOR, that just don't need change.

 

[crshears]

Late to the thread...

Quoting davidbeach: No. Lines can be reclosed, every other component of the electrical power system gets one chance; if it faults a lockout is tripped and somebody goes there to inspect, analyze, and evaluate.

To which I respond: it depends which "every other component" you're talking about...

By way of exception to the general rule posited, our utility has numerous outdoor low voltage [ 14 kV, 28 kV and 44 kV] busses; especially in the spring these are subject to trips due to animal incursion, mostly from raccoons but also by squirrels and the larger birds, especially crows and ravens. In many parts of the province, distances and therefore travel times for field staff can be considerable, which would lead to extended outages and customer interruptions while awaiting inspection. As a consequence our policy on outdoor low voltage busses is to allow ONE manual pick-up attempt following automatic removal from service; if the attempt is successful and all equipment is returned to service inspection is deferred to normal business hours Monday to Friday. Only if the attempt is unsuccessful do we promptly have "somebody go there to inspect, analyze, and evaluate."

Hope this helps.

CR

"As iron sharpens iron, so one person sharpens another." [Proverbs 27:17, NIV]