Arc Flash Hazard? Yes or No? 3

[FEinTX]

According to our arc flash study, we have a maximum of 47,000 A available 3-phase fault current coming into a 600 V cabinet via some really big copper buswork and wires. It connects to a main circuit breaker which is rated at 50,000 A. However, there are also three 4 AWG wires that tap off the line-side of each phase (no fuses) and go to an adjacent cabinet to a much smaller circuit breaker which is rated at only 35,000 A. The arc flash study is treating this all as a single electrical node, hence, the 47,000 A fault current exceeds the maximum current of the circuit breaker in the event of a bolted fault. Case closed. Or is it?

Question - Can #4 AWG wire carry 47,000 A of electricity for any meaningful period of time? If not, what would be the limit for this size wire?

Seems to me, the #4 wire would serve as a fuse link and simply vaporize with this kind of current, but this is way outside my training (my background is microelectronics). Any help is greatly appreciated.

Thanks,

FEinTX

 

[dpc]

The model can easily be expanded to include the #4 AWG wire in the network. Then the fault current at the smaller circuit breaker can be calculated.

47 kA will definitely melt the wire in time. This in itself is a major arc-flash hazard. But the breaker will probably open its contacts in a cycle or less. That's another problem if the breaker can see more current than it is rated for.

It's important not to confuse arc-flash hazards with short circuit rating issues. They are two different (but somewhat related)issues. The fact that all breakers are adequately rated for the expected short circuit current does not mean there is no arc-flash hazard in the equipment. The arc-flash hazard is based on an arcing fault inside the equipment. In fact, the arc-flash calculations assume that all breakers are adequately rated and operate as expected.

 

[TurbineGen]

The #4 wire will certainly act as a fuse. However there will be a temperature curve associated with that. It won't fail immediately. The melted wire presents the problem of melted metal which as dpc said is an arc flash hazard in itself. Also, this is NOT intended to clear the circuit. That should be done by the breaker. If your fault current exceeds the interruption rating of the breaker, you cannot rely on the wire to clear the circuit, even though in practice it might work.

------------------------------------------------------------------------
If it is broken, fix it. If it isn't broken, I'll soon fix that.

 

[ScottyUK]

When I read the OP there doesn't seem to be a breaker upstream of the tapping point for these 4 AWG cables, or at least not on the LV side of the transformer. The 4 AWG wire sounds like a slightly larger-than-normal ignition wire for an arc-flash test. Grossly undersized, it will probably snap due to the mechanical forces during the fault and initiate a very high energy arcing fault.


----------------------------------

If we learn from our mistakes I'm getting a great education!

 

[VTer]

Well,what is the available Isc at the adjacent cabinet with the CB that you refer to is being fed via the #4 AWG? If it’s above 35,000A then you have a problem not with the wire size but with the capability of your circuit breaker to interrupt and trip on a fault above its rated maximum.

"Throughout space there is energy. Is this energy static or kinetic! If static our hopes are in vain; if kinetic — and this we know it is, for certain — then it is a mere question of time when men will succeed in attaching their machinery to the very wheelwork of nature". – Nikola Tesla

 

[jghrist]

Unless this is a short tap, it sounds like it will not meet the requirements of NEC 240.21(B). If it is a short tap, then the fault current will not decrease much and the smaller breaker will not have adequate interrupting capacity.

As noted by others, none of this says anything about arc flash hazards.

 

[FEinTX]

Thanks for all the great input!

Yes, there are actually 2 issues that I'm looking for help with:

(1) Will the smaller CB actually see 47 KA coming across the #4 wire? If so, what happens? Will it not open? Will it explode?

(2) The arc flash incident energy for the line-side of the main CB is 48.5 cal/cm2 at 18". This makes energized access to this cabinet off-limits even with Class IV (40 cal) flash suit. This is OK because there is nothing needed in this cabinet from a service point of view.

The adjacent cabinet is a different story, however. There are many service items and troubleshooting points. Things such as small CB's for 110 V circuits (receptacles, lights), sensors, and more. If we also have 48.5 cal/cm2 in this cabinet, then we have a very difficult situation for operations teams in that we must de-power the pad-mount transformer for many routine tasks. If, however, we have < 40 cal/cm2, then we can at least open the door to this cabinet with a 40 cal flash suit and operate breakers and do some limited troubleshooting using switching sticks while maintaining a safe distance.


Note - we are posing these questions back to the company that performed the original arc flash study, but I'm very interested in what I can learn from those of you at this site.

Thanks!

FEinTX

 

[dpc]

The only way to know what the fault current can be at the smaller breaker is to calculate it. A certain length of #4 wire has a certain impedance. It will reduce the fault current. Impossible to say how much without more data.

The breaker does not "know" (in general) how much current it can interrupt, so if it sees current exceeding its trip setting it will trip and the contacts will open. What happens next depends on a lot of factors, but if the available fault current exceeds the breaker rating, the breaker can fail catastrophically, causing fires, bodily injury and other exciting things. It's also an NEC violation. However, this is probably still preferable to having the #4 wire melt, which very likely going to create a major arcing fault.

 

[Zogzog]

You can operate your breakers from outside the arc flash boundary using remote operators.

 

[TurbineGen]

1) Will the breaker see the 47kA? - Probably not, but you need to have provisions for worst case scenario. There may be significant impedance in the 4 AWG wires to reduce this to the CB's rating, but you must calculate this.

Will the breaker explode? - Again, not likely, but it could. It will likely take several more cycles to clear the fault, and get damaged in the process. Or it could just melt down. You never know. In the world of arc flash, you need to calculate for the worst case.

2) Perhaps the simplest solution to this problem would be to install a set of fuses. 60kA rated fuses are not that expensive. You could install them on the 4 AWG tap wires before the breaker.

------------------------------------------------------------------------
If it is broken, fix it. If it isn't broken, I'll soon fix that.

 

[rcw retired EE]

Using typical cable impedance data with the MVA method for calculating fault currents, and assuming that the wire impedance is similar to a cable, I calculate approximately 8 feet of #4 awg wire will drop the 3-phase fault level from 47 KA symmetrical to 35 kA symmetrical at 600 volts.

But at 35 kA the #4 awg is gone in less than 1 cycle. The fault current has to be below 20 kA to have a chance of saving the #4 awg wire. (See the Okonite nomograph.)

http://www.okonite.com/engineering/short-circuit-currents.html

If all of the assumptions are accurate, you may not have a problem with the breaker being overrated. A bolted fault on the downstream side of the breaker may not be greater than the 35kA breaker rating, if the melting of the wires doesn't affect the breaker's ability to clear the fault.

A fault on the breaker's line side terminals has to be cleared by the upstream protection; probably the transformer high voltage windings’ protection fuse or breaker. Clearing time will be a lot longer than the wire’s rating.

I would guess that your breaker is OK and your arc flash danger is off the map.

 

[rhatcher]

If a breaker sees a fault current that is greater than its' interrupting capacity, it may explode. What is certain is that regardless of how fast the breaker opens, the fault will not be interrupted simply by the contacts opening. The fault will remain until the circuit is (burned) opened enough to provide a large enough air gap to extinguish the arc at the zero crossing. Normally this kind of extended arc is accompanied by a very rapid expansion of superheated air (BOOM). This is a major arc-flash hazard. I would definitely question this discrepancy on the part of the arc flash survey.

However, from this forum there is a good soluion to the problem. TubineGen said:

"Perhaps the simplest solution to this problem would be to install a set of fuses. 60kA rated fuses are not that expensive. You could install them on the 4 AWG tap wires before the breaker. "

I agree and give a star for that suggestion.

 

[dpc]

The fuses solve the problem IF they are a tested combination with the small breaker.

 

[racobb]

On the assumption that the fault duty was lower when the equipment was installed:

I would replace the #4. It might have met and still meet the tap rule but needs replacing. You have not mentioned the breaker sizes so we do not know yet.

If the fault duty is above the rating of the breaker, then you either need to replace it or limit the current, as has been noted.

Once I see the fault duty exceeding 80% of the breaker rating, I start to get nervous. I sure can't handle the duty that I once could, but I think it is age and poor maintenance!%-) I have seen some equipment that I would hope would never need to interrupt anything near maximum rated!

"Will the smaller CB actually see 47 KA coming across the #4 wire? If so, what happens? Will it not open? Will it explode?"

If you cannot calculate the necessary currents and implement the appropriate mitigation, then please find someone to help you who can. With all due respect, you said your background is in micro-electronics and I tnink you have realized that this isn't an IC (no offense intended).

Alan

 

[rhatcher]

dpc, I think that I understand where you are coming from. There is a concern for using series connected devices to interrupt faults (ie. fuse before breaker or breaker before breaker). In the past, I understand that sometimes molded case breakers were connected one after the other in series with the idea that the interrupting rating could be added to equal, and interrupt, the total available fault current.

This is very questionable in my opinion and I would agree that it can only be counted on if the devices are tested for that capacity. I would say UL tested at a minumum.

However, I do not think that this rule would apply when the upstream device is, by itself, capable on interrupting the fault current. A CL fuse operates in less than one cycle and is, by definition, used to reduce the fault current contribution to downstream devices. At least that is what I believe. Is this wrong?

 

[davidbeach]

That's how it used to be thought, but various events have eliminated the acceptance of the up-over-down method of determining an effective let-thru current. The only acceptable method at this point is the use of listed, tested, combinations. Breakers are dynamic and may slow down CL fuses compared to how fast the fuse would interrupt in the absence of the circuit breaker.

 

[LionelHutz]

The fact that the 4awg could fail and cause a huge arc-flash fault in the busswork destroying all the busswork should be of as much concern as the fault rating of the breaker. I'd think a few corrective steps are in order.

I would install current limiting fuses at the bussbar. These fuses should almost be installed on the buss - in such a way that the bussbar to fuse seperation and wiring gives little chance the wires between the bussbar and fuses could short. These fuse will protect against a short circuit becoming a huge failure that blows up the wiring and the bussbar. With this protection, the arc flash rating of the tap cabinet should also be lowered to a level where work can be performed.

I would increase the wire size. I'd think about 1/0 wire is suitable for the application once current limiting fuses are installed. This wire also should have some decent bracing so it doesn't jump around during a fault - look at wire clamps or other such methods.

I would also change the circuit breaker in the tap cabinet to one rated for at least 65kA. You can not calculate the reduced current due to the fuses and assume a breaker of that current can be installed. You could install a long enough wire run that the current is reduced to match a lower rated breaker but once you upsize the wire this likely isn't a very practical solution.

 

[stevenal]

"Question - Can #4 AWG wire carry 47,000 A of electricity for any meaningful period of time? If not, what would be the limit for this size wire?"

t=(k*A/I)^2

k is determined by conductor or insulation. Bare copper is 0.0845, PVC is 0.03465.
A is the cross sectional area in circular mils. 4/0 is 211600.

Assuming bare copper you have 0.145 s before it fuses. Any type of insulation covering will fail much sooner.

Unless your conductor is supported like bare copper so that melting insulation will not cause a second fault; no.

 

[Zogzog]

Upgrading the breaker is easy, I can help you with that part.

 

[stevenal]

Sorry, I gave the A value for 4/0. #4 copper is 41740 cmil. It will fuse at 0.0056 s. Even a properly sized breaker will not prevent the second fault from occuring and a possible arc flash event.

How long is the #4 tap?