Buss differntial

[mbrooke]

Are their any special relaying considerations for main and transfer bus schemes where the main buss has a midpoint isolator (air break)? Or do I need a differential relay with dynamic zone selection? The goal is to be able to open the midpoint isolator and de-energize half the main buss for service if need be or bring the system back on line sooner if one section of the buss faulted.

 

[Bugman1400]

It would be best to have CTs at the isolator and break up the buss diff into two zones. However, in theory, as long as what goes into the diff zone is what comes out, you should not have an erroneous trip. If you de-energize part of the bus and its circuits then, you should not have any extra current flow that would cause an operation unless you start monkeying around with the CT secondary circuits of the de-energized circuits. Also, if the load current that is put on the xfer buss is appropriately accounted for by one of the feeder CTs then, you should be okay.

 

[JoeStillman]

The nice thing about differential protection is that it is inherently zone-selective. But with a main-tie-main arrangement, it's a good idea to have the branches selective with whatever is feeding them upstream. In that case, the 50/51's would be "dynamic" - sending their "wait-a-minute" signal to the tie or the other main.

 

[templdl]

Consider a pilot wire relay?

 

[mbrooke]

It would be best to have CTs at the isolator and break up the buss diff into two zones. However, in theory, as long as what goes into the diff zone is what comes out, you should not have an erroneous trip. If you de-energize part of the bus and its circuits then, you should not have any extra current flow that would cause an operation unless you start monkeying around with the CT secondary circuits of the de-energized circuits. Also, if the load current that is put on the xfer buss is appropriately accounted for by one of the feeder CTs then, you should be okay.

I could in theory get away with one duss differential zone since if all the breakers are tripped on one side and the buss mid point is opened, all the CTs will read zero on the tripped breakers. But if one of those breakers is serviced, that could cause the relaying to mis-operate?

 

[mbrooke]

Consider a pilot wire relay?

How would this work or be applied?

 

[Phil Corso]

Gentlemen...

I presume you are discussing a partial current-differential scheme! Partial is defined as CT's only in the source and bus-tie section and no connection with outgoing feeder CB's. If so, and if CT's are properly located, then removal of any one of the CB's from service should not jeopardize protection of the remaining circuits! Are the CT locations known?

BTW, zero-currents in CT's is not necessarily detrimental to proper operation of the protection scheme!

Regards, Phil Corso

 

[Bugman1400]

How would this work or be applied?

It wouldn't. Pilot wire relays haven't been used in years and have been replaced by much better microprocessor relays.

Also, as relay engineers will tell you, a oneline is worth a thousand words. Do you have one available with the proposed location of the isolation point, the mains, and the CT locales?

 

[ron]

Although probably way below the thought level of the posters here, I though I would just put a link for those of us needing a reference

https://selinc.com/api/download/102548?id=102548

 

[Phil Corso]

Thanks Ron...

The Crossover-over zone in Fig 7 of the ref you presented shows what I meant about insuring proper location of CT's!

Phil

 

[mbrooke]

It wouldn't. Pilot wire relays haven't been used in years and have been replaced by much better microprocessor relays.

Also, as relay engineers will tell you, a oneline is worth a thousand words. Do you have one available with the proposed location of the isolation point, the mains, and the CT locales?

My apologies for the delay :ashamed1:

 

[templdl]

It wouldn't. Pilot wire relays haven't been used in years and have been replaced by much better microprocessor relays.

Also, as relay engineers will tell you, a oneline is worth a thousand words. Do you have one available with the proposed location of the isolation point, the mains, and the CT locales?

Yes, the old telephone type wires haven't been as the means on coordination may be fiberoptics communications channel today.
And the microprocessor technology has been awesome at replacing the older electromechanical protective relays such as the infamous HCBs. The microprocessor has done wonders at improving protective relaying.

 

[templdl]

It wouldn't. Pilot wire relays haven't been used in years and have been replaced by much better microprocessor relays.

Also, as relay engineers will tell you, a oneline is worth a thousand words. Do you have one available with the proposed location of the isolation point, the mains, and the CT locales?

Yes, the old telephone type wires haven't been as the means on coordination may be fiberoptics communications channel today.
And the microprocessor technology has been awesome at replacing the older electromechanical protective relays such as the infamous HCBs. The microprocessor has done wonders at improving protective relaying.

 

[Bugman1400]

My apologies for the delay :ashamed1:

Typically, the Main-Tie-Main will have two buss zone with the Tie CT shared by both zones.

 

[mbrooke]

Typically, the Main-Tie-Main will have two buss zone with the Tie CT shared by both zones.

This ones is a tad different, at least for me. The tie is not automatic since its a air break disconnect. It is simply to sensationalize the buss after a buss fault and bring only half the circuits and one transformer in service while repairs are being made to the other. The buss protection must trip both sides for a buss fault, but hold when only half is energized. If two zones must be used, who gets the buss coupler?

 

[templdl]

My apologies for the delay :ashamed1:

I love your onerline example and symbols.

 

[topgone]

This ones is a tad different, at least for me. The tie is not automatic since its a air break disconnect. It is simply to sensationalize the buss after a buss fault and bring only half the circuits and one transformer in service while repairs are being made to the other. The buss protection must trip both sides for a buss fault, but hold when only half is energized. If two zones must be used, who gets the buss coupler?

Didn't you make sure your control logic short/open the respective CT outputs of those CTs of system elements that are to be cut out/switched unto the system. All CTs of all incoming on one side while outgoing CTs making the other side of the differential bus protection is the rule. What confuses me is when your bus ties are opened, splitting your main bus and whether or not your original bus differential protection still work!:?

 

[mbrooke]

Didn't you make sure your control logic short/open the respective CT outputs of those CTs of system elements that are to be cut out/switched unto the system. All CTs of all incoming on one side while outgoing CTs making the other side of the differential bus protection is the rule. What confuses me is when your bus ties are opened, splitting your main bus and whether or not your original bus differential protection still work!:?

Exactly what I am wondering. Every split buss scheme I have encountered is main-tie-main with a breaker run the middle (either open or closed) and 2 separate buss differential relays. In this case, one differential relay must operate correctly when the buss is intentionally split for service. I am not sure if its even possible.

 

[JoeStillman]

I would use three electronic 87 relays, one for each main and one for the tie. Then route all the CT's so that each CT secondary is wired to the three 87's in series. Add three A inputs to each 87 for the position of the mains and tie. That way, the summing calculation depends on what's being fed by what. Its all software.

If you had A switch inputs for all of the transfer switches, you could either ignore differential during transfer, or even do a more complicated sum.

 

[mbrooke]

I would use three electronic 87 relays, one for each main and one for the tie. Then route all the CT's so that each CT secondary is wired to the three 87's in series. Add three A inputs to each 87 for the position of the mains and tie. That way, the summing calculation depends on what's being fed by what. Its all software.

If you had A switch inputs for all of the transfer switches, you could either ignore differential during transfer, or even do a more complicated sum.

Can the tie point skip the CT (not the transfer buss coupler but the "hls")? The transfer buss will simply have a very low zero sequence pickup when not in use, and when in use carry either line or transformer protection.