Differential Protection in Switchyard 2

[Pitt03]

Currently reviewing the protection and control in a new 13.8kV outdoor switchyard and was curious to hear if others would recommended implementing differential protection as part of the switchyard protection scheme.

This new switchyard is fed from an existing Utility substation (12MVA transformer) via a re-closer in the substation and several hundred yards of overhead Arial cable. The new switchyard has a main bus with two reclosers that feed two secondary buses (separated by tie switch). The secondary bus then has (4) reclosers which will then be used for underground feeders to a industrial process on the site. All equipment and distribution in the new switchyard is pole mounted and uses overhead cable.

My question is weather or not it makes sense to use differential protection anywhere between the existing substation and the 4 new outgoing feeders out of the switchyard. I've seen bus differential schemes used in substations before where bus was used but am not sure if it would make sense with overhead cable? Not sure if CT's used in re-closers could be used as part of a differential scheme either?

I was thinking it may make sense to have some sort of line differential on the long overhead line between the existing substation and the new switchyard?

There is also local 13.8kV generation nearby On-Site which will be used to feed the new Switchyard in emergency situations. Generator will connect at the new Switchyard through a tie switch at the switchyard and will likely have an output breaker located close to the generator. Would it make sense to use differential protection anywhere between the generator and the switchyard for faults that the generator backup protection may be slow to react to?

 

[Marmite]

Protection is as much an art as a science, and the reliability you want to achieve is influenced heavily by how much money you want to spend. What you have described is a fairly cheap and nasty implementation of a substation, which is not uncommon in a rural lightly loaded network where reliability is not a key driver. Implementing full differential protection would be the equivalent of putting lipstick on a pig, but considerably more costly, and out of kilter with your overall design philosophy. You'd still be left with a pig.
Regards
Marmite

 

[Pitt03]

Marmite

Thanks. I'm curious in your mind what makes it a "cheap and nasty implementation"? I tend to agree with you but was curious what items jumped out in your mind as leading to that conclusion. Does it have to do with the fact that overhead cable is used instead of bus? Or does it have to do more with the system topology from a reliability standpoint?

My only concern for where differential may be needed is on the line from the existing On-Site generation that feeds into the switchyard through a Gang Operated Air Break Switch. Will the 51V or any overcurrent protection at the generators react fast enough for a fault in the Switchyard?

 

[Pitt03]

Can CT's inside of re-closures typically be used for other protection elements or are they limited to use with the re-closer protection only?

 

[PHovnanian]

What are the consequences of a fault on that overhead section? What will differential protection do to mitigate outage duration and/or repair costs? Is the 12MVA transformer already well protected for a through fault? What sort of faults do you expect on the several hundred yards of overhead?

 

[davidbeach]

Typically the CTs that come with reclosers aren't wired to other protection systems.

 

[jghrist]

If the reclosers have microprocessor relay/controls, you could use a fast bus trip scheme, sometimes referred to as a reverse-interlock scheme, to protect the bus. The recloser feeding the secondary bus would have a fast trip that is blocked from tripping if there is fault current sensed by the feeder recloser. For a bus fault, there is no blocking signal from the feeder recloser and the bus trips on the fast trip.

 

[DTR2011]

From what I've seen, reclosers typically have 1A nominal CT secondaries. It is possible to order relays with this rating (I think GE offers dual inputs as standard). I don't recall seeing the internal recloser CT's being shared with other protection devices. I have seen slip on CT's placed over reclosers for various reasons. Reclosers generally have an internal burden resistor across their output to allow for removing the 14-32 pin cable while the unit is in service, without opening the CT secondaries. I don't recall exactly, but I believe the last recloser I worked on had maybe C100 CT's. Maybe a Low Z diff scheme could be used.

The OP has not mentioned the specific recloser controls. In the USA, as far a recloser "controls" it is basically Cooper and SEL. Our local utility uses the fast bus trip scheme as the bus back up scheme, behind High Z Bus diff. This SEL fast bus scheme has been in place for 20+ years and its still installed on new substations. There is an excellent SEL applications guide (try the SEL 251 manual), which goes through the whole implementation. An added benefit of the SEL version of the scheme is the ability for the bus relay to back up a failed (or out of service) feeder relay and only trip the faulted feeder, not the whole bus. This makes any kind of work on the feeder relay possible (taking out of service for testing, settings changes, etc), while still having feeder protection. Certainly something that is missed from the days of E/M relays, where one could be taken out of service one at a time for service work.

Another consideration with trying to modify recloser controls, is that they are not wired like breaker controls. Trying to cut into the trip or close circuits is not as easy as with traditional breaker controls. With that in mind, perhaps the fast bus trip scheme implemented via a communications scheme (and internal logic) such as IEC 61850, SEL mirrored bits or Cooper Peer Comm may be the most economical way to go.

 

[Pitt03]

Yes re-closer CT's are 1000:1 CT's.

As I complete fault analysis I notice that there are potential issues with re-closer fault ratings and CT saturation due to their close proximity to a generation station. 23MVA generation station at 13.8kV is located approximately 500ft from this existing substation and new Switchyard.

Because of the high X/R ratio of the generators (aprox 30) the calculated momentary fault ratings seen by re-closers is greater than re-closers momentary current rating. Generation station fault contribution combined with contribution from existing 12MVA substation transformer provides about 11kA interrupting current and 19kA momentary current. Proposed re-closer has 14kA interrupting and 18kA momentary.

CT calculations also show that 1000:1 re-closer CT's will saturate quickly with fault current and high X/R ratio.

Has anyone had any experiences with re-closers located close to generation stations?

 

[ScottyUK]

Pitt03,

You've correctly identified a problem that many engineers overlook, which is the very high X/R ratio and the long time to the first current zero which are typically founbd close in to generating plant.

While I can't comment on reclosers, it isn't uncommon to find distribution circuit breakers pushed into service where a much larger, heavier and more expensive generator CB is needed. It isn't a problem until there's a fault close in to the generators, so many sites are unaware of just how bad things could get until their switchgear burns down. This design error is almost completely confined to small and medium-sized generation sites embedded on the distribution system: distribution-class breakers simply aren't big enough to handle the output currents of the large machines. These small sites are typically very cost-sensitive, and expensive luxuries like a generator CB aren't always provided.