Breaker Fail Philosophy

[electricnewbie]

I am fresh out of college and entered a Substation/Transmission Engineering position, where currently there is a debate on Breaker Fail philosophy.

The question is:
On a radial 115kv substation (designed to expand to breaker and a half scheme) is it best to:

1. Use a separate 50BF relay initiated by the primary and secondary line protection relays, or
2. Program in a breaker fail scheme into both the primary and secondary line protection relays eliminating the need for another relay, knowing that the breaker fail scheme is a Breaker Fail, not a Relay Fail scheme.

The argument seems to be between saving money by reducing relays (since programming into the relays does the same thing as having a separate 50BF relay), and making it easier to test the breaker fail by having a separate relay with separate test switches, etc.

Any suggestions or comments on the best breaker fail philosophy, or what is used in your communities/companies?

 

[davidbeach]

I see no advantages to a separate 50BF relay if every relay does its own breaker failure.

 

[rcw retired EE]

A single BF relay needs wiring from both the A & B relays for BF initiate signals. Wiring from the two independent and supposedly separate protection systems comes together on the BF relay possibly compromising the redundant and separate philosophy. A problem on that terminal strip could kill power on both protection systems, depending on the design.

Keeping the BF inside the A & B relays keeps the systems isolated.

 

[submonkey]

Derekv,

I prefer to use two separate breaker fail relays - one
on the A protection, and one on the B protection.

The breaker fail "initiate" input is energised by
the main protection. The trip output of the BF relay
takes its positive from the same point as the
"initiate" input.

The arrangement is designed to ensure that a busbar
trip can be issued only if both of the following occur:

(a) The main protection has operated.
(b) The BF sees current after the CB should have opened.

The scheme is more secure than one programmed into the
main protection relay, because two separate relays
must "vote" together before a bus trip occurs.

If, due to a programming error or malfunction, the
main protection relay operates incorrectly, the worst
that can result is the tripping of the protected object.
No bus trip can result because the BF relay will
perform its own current check.

If the BF relay operates incorrectly, the bus cannot
be tripped because the main protection has not provided
a positive for the BF relay to trip the bus with.

An added advantage is that the wiring and settings for
the BF relay are simple, standard, and well understood
by the field staff.

Bear in mind that testing a BF scheme for security is
very difficult. Typical relay tests involve providing
some input and checking that the right response occurs.
Proving that a scheme will NOT trip in all circumstances
when it shouldn't is practically impossible.

Some will argue that my approach is too costly. This cost
should be weighed against the cost of an incorrect busbar
trip. In a radially-fed substation, a bus trip may not
be a big deal.

If two separate BF relays cannot be provided, I would
prefer to program an internal BF into each main relay
rather than crossing A/B wiring.

Thanks,
Submonkey

 

[davidbeach]

If the main relays are capable of breaker failure protection, using separate BF relays only adds complexity and additional points of failure. That, plus requiring more rack space. Using the main relays for BF means that BFI is all in logic. Using separate relays means that you have a BFI bus in addition to the trip bus, or you have diodes in the trip circuit so that protective trips trigger BFI but control switch operations don't. We've been through the analysis and have found that our security and dependability have both increased as we have transitioned from separate BF relays to integrating BF into the main relays. When using separate BF relays, there would be one per breaker. Using the main relays, we can have up to four relays doing breaker failure (though not necessarily all at the same time) for an increase in reliability and redundancy.

 

[marks1080]

Go with option 2. Option 1 is from an older school of thought when current technologies didn't exist. To install 50BF relays now is really a waste of time and money for no benefit.

 

[electricnewbie]

Thank you all for your suggestions. Hopefully we can come to a conclusion on how we will apply BF for future applications.

 

[bacon4life]

Another less dependable option is only having the BF logic inside of the primary line relay. With breaker and a half you need the BF relay to see which breaker is stuck. Our typical package has an SEL421 and and an SEL311L relay. However the 300 series only has one CT input and they don't have a 400 series line differential relay. It isn't ideal, but it is simpler than adding on an independant BF relay to the 311L.

If we had dual batteries, I would be more likely to want truely independant and redundant protection. With only a single battery, the remote backup requirements for battery failure seems comparable to a breaker failure with a primary relay failure.

 

[dfdt]

becon4life
I am told that SEL is about to release SEL411L relay which has dual CT input and will replace SEL 311L relay

 

[bacon4life]

This is a great summary of several different ways to do BF with microprocessor based relays.

http://www.pes-psrc.org/Reports/Ancillary_PC_Functions_Common_to_Multiple_Protective_Relays_10.1.pdf

 

[electricnewbie]

Would it be easier to do BF on all tripping device, (differential lockouts, primary & secondary line relays, etc) though a dedicated BF relay?

Thank you bacon4life for that document.

 

[davidbeach]

No, that means another relay. The protective relays that can do their own breaker failure tripping can also have BFI inputs from other devices that can't do breaker failure.

 

[electricnewbie]

Good point. There is a large debate for old school vs new school here, and you all have great information for me to have an understanding in the debate. Thanks again.