SEL Mirrored Bits

[PeterVenkman]

I am wondering if anyone has daisy-chained a bunch of relays together using mirrored bits to carry out zone blocking to reduce bus arc flash levels? Sending a bit from one relay to the next until it is sent up to the relay bus relay. I am wondering if it is faster while also providing the added benefit of letting you know if communications have failed.

About how long does it take for a mirrored bit to transfer from one relay to another?

About how fast are high speed and non-speed physical contacts?

 

[cranky108]

Why not use a mirrored bit processor? And to the point, what happens if you take a relay out of service? Do you loose your arc-flash protection?

I personally avoid the whole issue and use a bus differential.

 

[PeterVenkman]

It was more making due with whats there and just wondering if it is possible. Looking through the manual, it is slower than just paralleling physical contacts from the relays.

 

[cranky108]

If you want the relay tech.'s on your side, make them wear the moon suits to a few locations. They will typically do any idea that avoids that.

But to the point, fixing old gear is never easy. Our guys were all with some scheme like you are proposing, until we informad them that live gear testing required the moon suit anyway because they were taking the high speed system out of service to test each feeder, and they required an outage anyway to test the bus scheme.
This was versis the testing of each feeder in normal clothes, and taking an outage to test the bus differential.

The only way such a voteing scheme makes since is if you have guys in the gear on a regular bases, other than testing. Which in a substation is not typical.

 

[dpc]

MIrrored bits are very fast, but generally, you need to have multiple feeder relays sending restraining signals to a single main breaker relay. Mirrored bits is going to be point-to-point unless you use some kind of processor. We generally just use output contacts from the feeder relays and connect in parallel to a single input in the Main Breaker to use as torque control.

The hard-wired contacts are not as fast, but they will be still be much faster than waiting on the normal coordinated time overcurrent function.

SEL calls this "fast bus tripping" - they used to have some white papers on their website. We call it the poor man's bus differential relay.

 

[jghrist]

We used an RTAC to receive outputs from several feeder relays for a fast bus trip scheme and then sent a blocking signal to an SEL-351S in another substation over fiber with mirrored bits for blocking (no main breaker in the sub with the feeders). You need to be careful that the processing interval on the RTAC is set low enough. We initially had the processing interval at 100ms and the blocking signal was delayed too much, causing a trip of the station for a feeder fault. We got the problem resolved and have had no problems since.