Vacuum Circuit Breakers and transformer protection

[jado85]

I'm looking into installing a Vacuum Circuit Breaker(VCB) on a Switch gear that supplies a 3000KVA transformer, this transformer feeds 3 panels rated 800A in each of 2 500MCM parallel runs( see the attached diagram)

If I want to protect the secondary of the transformers I would need to have to monitor the current going to each panel(3CT's for each panel total 9 CT's), at least this is what I think should be done to protect the cables. my logic, the cables for each panel is rated at 800A and the available current from the XF is 3600A, if I have a fault between the XF and the panel I want to open my VCB when the current exceed the 800A ampacity for the cable.

I'm being told that monitoring the overall current on each phase(3 CT's on the secondary side) will provide the protection needed, how would that work?

 

[jim dungar]

I'm being told that monitoring the overall current on each phase(3 CT's on the secondary side) will provide the protection needed, how would that work?

Monitoring the overall current will provide overall transformer secondary protection.
Monitoring each individual feeder circuit will provide individual circuit conductor protection.

 

[zog]

Before you use a VCB I would do some research on prestrike effects from VCB's on transformers. Also, not using a main breaker on the secondary side is going to cause you some pain when you do your arc flash analysis.

 

[big john]

Before you use a VCB I would do some research on prestrike effects from VCB's on transformers. Also, not using a main breaker on the secondary side is going to cause you some pain when you do your arc flash analysis.

Definitely.

And unless OP has an unusual application where you want electrical remote-control of all this equipment, I'm not sure why you'd install a breaker at all.

Choose a fused air-switch it will be far simpler, cheaper, and more reliable.

 

[templdl]

Definitely.

And unless OP has an unusual application where you want electrical remote-control of all this equipment, I'm not sure why you'd install a breaker at all.

Choose a fused air-switch it will be far simpler, cheaper, and more reliable.

I have sold plenty of the Eaton type WLI load interrupter switches for this allocation. I'm not so sure CLE fuses would work with a 3000kva but you may need to go with an RBA expulsion fuse for the voltage you have specified, or did the OP say? I would take a wild ass guess that it may be 12470, 13,200, 13,800 or there abouts.

 

[jado85]

we have a fused air switch, the VCB will be a retrofit. The current installation does not meet code regarding the length of secondary inside the building and I don't have a location to add a main CB within 25 ft to meet code.

*each panel has a Main CB but like I said it doesn't meet code on location of the CB and that's why I'm considering the VCB.

Jim,
That's what I needed to confirm, I'm not crazy! Thank you.

 

[jado85]

Before you use a VCB I would do some research on prestrike effects from VCB's on transformers. Also, not using a main breaker on the secondary side is going to cause you some pain when you do your arc flash analysis.

Thank you for the tip, I've done some reading on prestrike effects. I can't say that I understand it fully.
In our case, we would switch off VCB once a year for maintenance, and obviously in case of a fault, how would the prestrike affects the life of the transformer and the VCB life?

 

[Tony S]

Similar situation to yours. I fitted CT’s to the transformer output for O/L and REF, the relay simultaneously opened the MV VCB and all the LV ACB’s. It was a bit more complicated than that but in essence it was a “quick fix”.

It was an unusual situation that I inherited and I can’t understand why a main LV ACB and feeder panel hadn’t been installed.

 

[zog]

Thank you for the tip, I've done some reading on prestrike effects. I can't say that I understand it fully.
In our case, we would switch off VCB once a year for maintenance, and obviously in case of a fault, how would the prestrike affects the life of the transformer and the VCB life?

Once a year I don't think will be a problem.

 

[Haji]

VCB may cause voltage surge whìle switching. May damage your transformer if it is dry type.

 

[jado85]

Monitoring the overall current will provide overall transformer secondary protection.
Monitoring each individual feeder circuit will provide individual circuit conductor protection.

In the case that we have 2 or 3 parallel runs I'm being told that one CT per phase regardless of the number of parallel runs feeding the panel would work, how does a fault on one parallel run affect the current on the other runs, what current would the CT see ?
example, one panel fed by 500mcm in 2 parallel runs,that's 2x400A ampacity (old standard), I get a fault on Phase A on one of the runs(Fault is line side of the downstream CB), and my CT is around 2 phase A runs, relay is set to trip at 800A, assuming that the constant load is very low there would be a chance that the majority of the fault current will go through the faulted phase and the relay will not trip untill the current in both runs exceeds 800A, but the faulted cable is only rated for 400A leading to damage/fire

VCB may cause voltage surge whìle switching. May damage your transformer if it is dry type.

Our transformers are Oil filled.

Once a year I don't think will be a problem.

Is this related to in rush currents?

 

[GoldDigger]

I see this as similar to the case of motor overloads protecting the wires from overload while the OCPD protects only against shorts and ground faults.
A shirt in one conductor can be expected to draw enough current to exceed the combined trip current of the breaker while any loser level overcurrent at the load will still split equally between conductors and eventually hit the thermal trip without conductor damage.
I would be more concerned about one of the parallel conductors opening without that event being detected. If you were required to protect against that you would need one CT and detector per wire.

 

[paulengr]

Before you use a VCB I would do some research on prestrike effects from VCB's on transformers. Also, not using a main breaker on the secondary side is going to cause you some pain when you do your arc flash analysis.

Prestrike was a problem years ago and so is restrike. They softened the opening as a result. "Modern" VCB's (those made in the last 20 years) don't do it. Check with Toshiba that made the original VCB's. Or CBS Sales. There are only 3 vacuum interrupter manufacturers in the world but dozens of vacuum circuit breaker manufacturers. I've used VCB's for years without prestrike issues.

There is a simpler way to get a main CB with that arrangement with lower arc flash, cheaper breakers, etc. Use bushing CT's on the secondary terminals of the tra nsformer to drive 50/51 relaying but trip the primary side CB. This is called a virtual main. Unlike a real 'main', the incident energy is much less and the window where incident energy is sky high on the secondary side is effectively zero. I like SEL so you could use either 2 751A's for this arrangement or a single 651. Also if you add 3 PT's on the secondary side for troubleshooting and feed them to the relaying you can do 87 relaying for free which even detects internal ground faults in the transformer. If its over 1000 V you need this anyways to measure voltages without exotic test equipment.

Fused disconnects at one time were cheaper but you can get VCB's in metal enclosed switchgear for close to the same price as long as its under around 600-800 A. Above that point due to physics you get multi-electrode VCB's and the price doubles with each doubling of the current limit. Price also goes nuts above 12.5 ka and again above 20 ka fault current, driving you into a fused solution or often a VCB with backup fuses which earns 60-100 ka ratings.

You may not be able to avoid secondary protection on the feeders. If the secondary cabling is rated 800x3=2400 A then fine. But if its only 800 A rating you might be able to get away with the tap rules if you have 800 A over current protection on each phase on the downstream side and you have at least 2400 A capacity on the bus they are tapped off of.

 

[jim dungar]

In the case that we have 2 or 3 parallel runs I'm being told that one CT per phase regardless of the number of parallel runs feeding the panel would work, how does a fault on one parallel run affect the current on the other runs, what current would the CT see ?

A single CT cannot protect the individual conductors of a parallel run.
If you are concerned with the individual conductor protection, you should look into 'cable protectors', which are special fuses.

 

[Tony S]

Summation O/C relays and C/T’s are an option in this situation.

 

[jado85]

Summation O/C relays and C/T’s are an option in this situation.

This might a cheaper solution than my $60K retrofit, what are the limitations/rules on installing those in the Xf low voltage cabin?

 

[Tony S]

This is where we have a slight problem. The companies I use are obviously UK based.

Even so, have a look for D. K. Moriarty. (Nothing to do with Sherlock Holms)
As an international company they will have a US base that will help.

It is a problematic situation you find yourself in.

 

[jado85]

A single CT cannot protect the individual conductors of a parallel run.
If you are concerned with the individual conductor protection, you should look into 'cable protectors', which are special fuses.

Besides overload situations, would a fault always be greater than 3600A(my transformer is 3000KVA)?
I know what the available short current is, is there a minimum fault current calculation?

 

[paulengr]

On shorts, you have line-line, line-ground, line a ground a line, and line-line-line. For medium voltage arcing current is close to bolted faults so that simplifies things. Once we get away from "faults", we get into overload/over current scenarios where we really just look at ampacity. In the world between these two scenarios outside of various unusual cases, the biggest concern is known as a high resistance fault which can fall through the cracks if you aren't using time a current a curves to design and set over current protection. The best defense against these type of faults is to useresistance grounding since they are almost always line-ground. As an example I've worked on 7200 V systems where the maximum ground fault current is 25 A theoretically with a dead short but trips in a couple seconds at 10 A. A neighbor recently "tested" the relaying in June by accidently dropping a ground cluster in an inadvertently energized box. Injuries are not pleasant but he walked away under his own power. 5 years ago another contractor "tested" one with a wrench to a PT at 4160 and walked away with relatively minor injuries (a couple red marks).