similar distance relay for main and backup protection 1

[144x]

Hi,a vendor has offered me 2 similar distance relays for 230 kv transmission line protection.usually we reqiure 2 different hardwares or 2 different algorithms in relays for distance protection.but this vendor has no such thing.I want to know your experience and ideas about using similar distance relays for high voltage transmission lines .any comments would be appreciated.
thanx

 

[521AB]

What do you want to do?
Increase the probability to trip (increase availability)? Two different hardware and two different softwares (even if from same company). If relay "A" fails to trip for that fault, probably "B" will not fail there.
But you also increase the probability of "unwanted tripping", "A" can trip unwanted or "B" can trip unwanted (you decrease security).

If you want to increase security, "A" and "B" equals is probably the best,but you decrease unavailability.

Till 10 years ago everybody was on choise 1, wich now is under questioning by some people.

 

[slavagar]

Hi.
You have several sides for this issue:
1. Two similar relays are not good solution( you can ask ABB, Siemens, Areva, sorry , I have experience only with Eropean companies).
2. For my opiniom, two distance protection also are not good solution, but it's only opinion.
3. I have good experience with two type of protection, first is distance ( include directional earth fault protection) and second is line differential.
4. But if you used only distance protection , is better use two types of relays or two different companies.
What is a problem in this solution:
1. Two types of SW and HW.
2. Two types of communication cables
3. Two types of spare parts.
Before sveral years we "found bug" in some good distance relay ( in 1% of fault (very special faults) it's work uncorrect)but second differential relay work correct.
I have good expirence for generator-trafo protection, today we put only different relays, also for full redandancy and also for overlapping principle.
What about autoreclosere function: you used other relay, put in one of distance relay, put in both of them and use master - follower logic.
Are You have different cores from CT and VT?
Are you have DC1 and DC2 for relays power supply and trip coils?
Are you have BBP+BFP?
See, all small things are influence on your choice, sorry, customer choice.
521AB also put some dillema, what you prefer? unwanted trip or not trip in fault time?
This is our (protection engineer) bigest problem, and I think we don't have answers.

 

[dpc]

Use of identical distance relays is fairly common in the US. The only potential risk is some flaw in the relay's algorithm or A/D conversion that would cause both relays to fail to trip when required. Unless you're talking about a 2 out of 2 scheme, a false trip in either relay is still a trip.

There is no perfect solution.

This question once was raised in a meeting with Dr. Ed Schweitzer. Of course, he saw nothing wrong with two identical relays as long as they were both SEL When asked about the risk of a failure to trip due to relay firmware problems, he simply said "That has never happened with SEL relays". Since SEL is one of the few relay vendors that actually inform customers of problems with their relays, I tend to believe him. But there will always be some risk - like most things in life.

 

[Hoxton]

Some while ago I worked for a UK protection relay manufacturer. At the time, the CEGB (now National Grid) seemed to protect each 400kV line with two different types of relay (distance / differential) from different manufacturers. I do not know what the practice was at lower voltages. Their philosophy was a "1 out of 2" to trip. They were thinking of a "2 out of 3" voting system, but it seemed to be too expensive for the possible benefits in reliability.

I did note the other year that there was a major outage, shutting down part of London, due to a false trip, because someoneone had installed a 1 amp relay instead of a 5 amp relay at a major substation.

 

[slavagar]

What we need and what we want : decrease risk of unwanted (false) trip or decrease risk of no trip.
Second: of course cost, differntial protection is request also PLC or FO connection between two ends ( several ends more problems).
Maybe in several cases possible use only one relay, NB line provide backup protection with 2-d or OR zone.
I'm also remamber now, in one substation 161kV we use two indetical dist protection from 1999 w/o any problem,.

 

[stevenal]

From WECC:

"At least two relay systems are required to achieve dependability for transmission line protection based on NERC/WECC Planning Standards. Each system shall operate from separate AC current and potential secondary windings, have separately fused DC control voltages, and should have dual circuit breaker trip coils operated by separately fused DC sources. The redundant systems should be designed to eliminate the possibility that a single protection system component failure can disable all applied systems."

They stress eliminating common components, not algorithms or design.

 

[davidbeach]

stevenal, could you please cite where that requirement is documented?

 

[rcw retired EE]

Stevenal and Davidbeach - this same language is also in the Ontario Transmission System Code and the NEPCC (North East Power Coordinating Council) reliability requirements that apply to all systems 230 kV and above.

dpc- For several years I never used SEL relays because I was in grad school at WSU with Ed Schweitzer and thought I knew the student engineers who were assembling the SEL121 relays. I had too many beers with them to think I could trust their solder jobs. I was wrong.

Where we have used SEL relays as backup or alternate relays on dual protection systems, the SEL relay usually clears the fault first.

 

[stevenal]

http://www.wecc.biz/documents/2006/Standards/Guide_for_Specs_for_345kV_relaying-Final-RevB.pdf

 

[stevenal]

davidbeach,

Also check your WPRC '06 proceedings for a presentation by D. Angell. The Power point makes the same recommendation. Note while they advocate no single failure point, separate fuses off the same battery/charger system seems to be enough. A single open circuited cell might take out both systems. And separate trip coils eventually operate the same mechanics.

 

[davidbeach]

stevenal, thanks. The 345kV requirement cleared up my confusion.

 

[lz5pl]

According to my experience two protections from different manufacturers is better solution than two from the same type (even it is SEL ).
I had a case in 220kV line in Egypt, protected by one 7SA513 and one REL521. Ground fault happened around 2,5 km from the substation (totally line was about 43 km). ABB relay triped, but Siemens relay doesn't. As the station was still under guarantee the customer put in a claim for defective relay. When uploaded fault record from REL521 I found that actually the fault have been highly resistive, probably leakage over the surface of an insulator string. Due to the specific of it's polygonal characteristic 7SA513 doesn't see it as a fault, but as a heavy load. If both relays were from this type fault could not be cleared before to evolve to more serious. Good point was that setting values were calculated by the customer, so the claim was lifted after my explanation.

------------------------
It may be like this in theory and practice, but in real life it is completely different.
The favourite sentence of my army sergeant

 

[521AB]

lz5pl

If you had 2 ABB relays, they had both tripped. By using 2 different relays, yu are increasing the complexity for the customer (he has to read and understand 2 different relays, read 2 manuals, attend two different courses...). I also don't know actually what is better.
Buy using ABB and Siemens the customer has increased its probability to fail. MAIN1 and MAIN2 should behave in the same way. What had happened if the ABB relay had been out of service (for maintenance for instance?).

Regarding the customer claim: I woul be usually very careful before blaiming the relay. First I would request a post-fault analysis to the customer (even by paying some money for that), after that I maybe have to see my setting strategy... :-(

Another comment: I have a lot of colleagues who believe in the "perfect" relay.
But I think I will put a new thread for that..