MV Metalclad Bus Differential 2

[gordonl]

I'm in the process of replacing an old 15kV metalclad lineup and am wondering what others are using for bus protection. In the past we've used graded overcurrents which gave delays of upto several seconds for solid bus faults. I'm currently looking at blocking & tripping of instantaneous elements or bus differential and would appreciate any comments on current practices and experience.

I'm leaning to the blocking method right now because of the low cost of implementation. The sub is double ended with the tie breaker normally open. All loads are cable feeds to transformers.

 

[peterb]

Gordonl -
By the blocking method, I presume that you are talking about the scheme where the main breaker instantaneous element is blocked by operation of any of the feeder instantaneous elements. This provides a relatively simple and certainly inexpensive means of bus protection. My reservation in using this scheme is the possibility of mis-operation due to wiring or external relay fault/failure - what if say the power supply fails on the feeder relay on a faulted circuit?
Depending on the importance of the bus and the critical clearing time, I would usually specify a high impedance bus differential scheme. Yes, you need extra CTs, but the added security provides a lot of peace of mind.
On the other hand, as you are coming from a situation where time graded overcurrent provided the bus protection, fast operation probably isn't that important to you and the interlocked zone-selective scheme, with overcurrent backup, is probably perfectly adequate.

 

[busbar]

I thought that bus differential relaying should be strictly high-speed (having no intentional time delay) If installed correctly it should be very secure (not prone to false trips). If possible, get a copy of: ANSI/IEEE C37.97-1979 Guide for Protective Relay Applications to Power System Buses A *description* of this document is at:

http://standards.ieee.org/reading/ieee/std_public/description/relaying/C37.97-1979_desc.html

 

[chcastro]

There's a very interesting article in the Transmission & Distribution World magazine, Dec 2001 on high-speed bus protection. The design requires 100ms clearing to avoid a serious safety hazard and extensive damage to the switchgear, particularly when adequate arc containment or venting was not provided. It describes the optical and overcurrent scheme used.

 

[dpc]

gordonl,

For critical applications, we would recommend the high-impedance bus differential relaying. This is very reliable and high-speed. I believe Schweitzer has a new digital relay out for this function.

If you have newer digital relays, you could consider using the "instantaneous" overcurrent function with enough of a definite time-delay to allow the feeder breaker instantaneous to operate. This is a compromise that is fairly easy to implement and should allow tripping much faster than several seconds. Because all the feeders serve transformers, this should be quite feasible, and eliminates need for interlocking.

Hope this helps.

dpc

 

[gordonl]

In the form of a limited survey, for industrial application metalclad line-ups have/haven't you used bus differential. If so for how long, if not anything instead.

Thank You for present and future input,
Gord

 

[dpc]

Gordon,

For metal-clad switchgear, we have recommended it on only two or three occasions, to deal with specific coordination issues and reliability concerns. These buses connected to industrial generation as well as typical loads. I would have to say that I have not seen it used very often, even for critical systems such as station service for large power plants. For outdoor substations it is more common.

For the situation where all of the feeders go to step-down transformers, it might be hard to justify the increased cost and complexity of bus diff. You can set instantaneous on each feeder relay so you really only increase protection for the bus itself and the risk of bus fault inside the metal-clad is low.

Just my $0.02.

 

[ongchazira]

I am not very clear about this blocking arrangement of instantaneous relays. I understand that for any fault whether it occurs on any of the transformer feeders or in the 15kV bus, the fault level will remain same. By using instantaneous elements how we are going to achieve co-ordination among the relays. It may so happen that the instaneous relay on the incoming feeder may operate first. In that case how the blocking arrangement is going to work? Are you combining it with some delays?. But I would prefer to use bus differential and I feel the blocking arrangement will become too complex if the number of outgoing feeders are more. As peterb has pointed out I think the critical clearing time has got a role to play. We had gone for bus differential schemes in some of substations to achieve better coordination as the critical clearing time was very less.

 

[gordonl]

In the blocking scheme the feeder has a contact which picksup when it sees a SC, this contact wires into the main incoming relay to prevent it from tripping. The main incomming would have a small delay to allow the contact to pickup. The main incomming would also have a time delay co-ordinated unblocked OC for backup in case of breaker failure, or false blocking. The cost of implementation is very low with electronic relays because intercell wiring is the only addition.

I've never used bus diff, what would occur if there was a failure of the intercell wiring for the CT's? Are the new relays smart enough to know the difference from a fault, or trip. If trip then I don't see how it would be more reliable. With the blocking method you would need a protection failure and a fault for system failure. With the blocking method I would detect a protection system alarm when a problem occured and could rectify without a trip occuring. (PS. blocking would occur only intercell, not inter sub, which I have read about but am leary to try)

 

[peterb]

Gordonl -
For what it's worth, nearly 100% of the many 15kV metalclad switchgear busses that I have dealt with over the last 30+ years have been equipped with high impedance differential protection. This includes chemical plants, utility plant auxiliary systems and cogeneration plants, in North America and internationally.
I realise that this is the opposite of what dpc has seen, but the major considerations must be the level of damage that can result from delayed fault clearance and the effect on the electrical system response. Although rare, bus faults DO happen in metalclad gear and the consequences can be catastrophic. I'm a believer, as I have seen the bus differential catch low-level faults that allowed repair of the bus insulation before major damage occurred, saving untold dollars and time.

For ongchazira - yes, the interlocked scheme does have some time delay built in to verify whether or not any of the downstream feeder instantaneous relays have operated before allowing the main breaker instantaneous to trip.

 

[peterb]

I just surfed into this link, which may be of interest -

http://industryclick.com/magazinear...aseid=9724&magazinearticleid=137056&siteid=30

 

[dpc]

Well, I'm not saying it's a bad idea, just that I haven't seen it done much. No question it improves protection and simplifies coordination.

But I do believe that bus faults in metal-clad switchgear must rank well below other failure modes in medium-voltage systems.

dpc

 

[gordonl]

Thank You All

As an aside, ABB offers light sensing protection on their metalclad lineups as well.

 

[busbar]

No new news here: it goes to protection philosophy. If you’ve had significant damage to switchgear that bus-dif protection may have saved, you tend to favor it; otherwise, no. One impossible-to-coordinate situation in industrial settings is 50 devices on a transformer-secondary breaker and downstream feeder breakers. And, there is often transformer percentage-differential relaying but that zone doesn’t usually cover the whole lowside bus. I think some figure that high-speed bus-differential relaying makes up for this ‘hole’ in protection.