True or False Current limiting? 2

[tulum]

I thought if you wanted to decrease the available fault current on a system, you could place the appropriate current limiting fuse ahead of it to limit the let-thru (this is what I have been doing).

Recently I was told that you can not do this without using a manufacturers published series rating chart.

True or False?

Regards,
TULUM

 

[dpc]

True - at least from a code/regulatory point of view. You must use a tested combination of upstream and downstream devices in order to take credit for any current-limiting. The US circuit breaker manufacturers can provide lists of these tested combinations, similar to their series ratings of circuit breakers.

The interaction between an upstream fuse and a downstream circuit breaker during fault conditions is dynamic and complex. The old "up-over-down" method we used to use for applying current-limiting fuses has been shown to be unreliable. The circuit breaker itself has substantial impedance during faults and sometimes the actual fault is much less than expected and the fuse doesn't operate quickly enough to limit current, but the breaker is still over-dutied.

Also, keep in mind that for a fuse to be considered current-limiting, it must clear the fault within the first half-cycle of the fault. If the prospective fault current indicates a slower operation than this, the fuse isn't providing current-limiting, at least per UL's defintion.

 

[resqcapt19]

Prior to the 2005 code, only listed and test combinations of "series" systems were permitted. When an active upstream device protects a downstream device with an AIC rating that is less than the available fault current you have a series rated system. Under the new wording in the 2005 NEC, an licensed professional engineer with experience in power systems is permitted to design a "series" system using upstream current limiting devices to protect the downstream devices.
Don

 

[dpc]

Thanks for the update, Don. I guess I need to get that new Code book ordered.

dpc

 

[davidbeach]

I believe that the permission in the 2005 NEC for calculated series ratings is a disaster in the making. If the breaker manufacturer's cannot reliably calculate performance of devices in series (all series listings are established by testing), how is an engineer with a whole lot less knowledge of the breaker in question supposed to calculate the rating? I, personally, would refuse to even begin to attempt to do so though I fall into the category of professional engineer the code would allow to do so.

 

[tulum]

Thanks for all the help guys - especially the code update.

However, I have only applied current limiting fuses for the purpose of decreasing fault levels on static systems such as cables and bus. Dynamic systems, such as breakers are another story all together.

Regards,
TULUM

 

[davidbeach]

tulum,

One thing to watch for, the current limiting is based on the fault current, at the fault, not the prospective fault current at the fuse location. So if you have a fuse applied at a location where there might be 60kA available, and the fuse is current limiting for currents above 50kA, then if you have a circuit the due to impedance limitations will only draw 48kA, there will be no current limiting by the fuse. The fuse may well still blow, but not fast enough to count as current limiting.

 

[resqcapt19]

The main reason for the code change is to provide a way to bring existing installed equipment into compliance. The building service and other electrical equipment may have had the correct AIC rating at the time of installation, but when the utility upgrades the distribution system and the available fault current is increased to a level higher than the rating of the electrical equipment, there was no way of correcting this under the 2002 code without replacing the equipment. Note that this issue was heavily fought out over 2 or 3 code cycles. The major players were the fuse and the breaker manufactures.

 

[alehman]

A big problem with the new provision is that it must be applied only with circuit breakers that will not open before the fuse has ample time to clear the fault. This rules out all current limiting breakers and many other molded case breakers that are are fast enough that they might open before the fault is cleared.

I concur with David. I'm a PE, but I wouldn't want to approve such an arrangement without a lot of careful analysis.

 

[mc5w]

You should NOT have a problem with 200 amp class J fuses that are installed ahead of 100 amp or smaller circuit breakers. SquareD and similar circuit breakers that have a 1/2 to 1 cycle clear time will coordinate with 200 amp class J time delay fuses for short circuits close to a panelboard involving a 20 amp circuit using #10 copper wire. A forklift truck driver shearing off a 40 amp power feed on 277Y480 volts WILL blow the fuses. This is actual plant experience.

Since the vast majority of short circuits are at the far end of the branch circuit the breaker will open the fault before the fuse does. For those few shorts that are close to the panelboard you would want the panelboard to be shut of by the feeder overcurrent device anyways.

The better circuit breakers such as SquareD and General Electric have a 1/2 to 1 cycle clear time for any short that is 1000% of the overload rating of the circuit breaker. This works well with 200 amp class J fuses. A 2 or more cycle clear time is going to have coordination problems.

There are fewer series ratings for 400 amp class J fuses but it turns out that 200 amp equipment is 1/3 the price of 400 amp. A 400 amp 4 wire wye feeder that is derated because of harmonic in the neutral is really a pair of 200 amp lines in parallel. Makes more sense to wire as a pair of 200 amp lines feeding 2 panelboards.

For shorts that are above the interrupting rating of the breaker the class J fuse of correct size will HELP the circuit breaker open the fault. The key is that the let through of the fuse has to be less than the let through of a 1/2 to 1 cycle clear time breaker at maximum rated intrrupting level.

For any circuit breaker that is still being made you should be able to get a UL listed series rating from the manufacturer. If your breakers are no longer made you have a Spare Parts Conspiracy problem anyways.

 

[alehman]

mc: Your statements pertaining to series combinations are true only if the breaker does not open within 1/2 cycle. The up-over-down method can be applied. Problem is many MCCB's open in less than 1/2 cycle at high fault current.

UL listed fuse/breaker series ratings have been permitted by the NEC for many years. I don't have a problem with using those listed combinations. The new 2005 provision allows a PE to evaluate and use UNLISTED and untested series combinations. That can be dangerous. I wouldn't do it unless I was very confident of the characteristics of the devices involved.

 

[davidbeach]

alehman,

The only breakers I know of that might be worth considering for "evaluation" are old power breakers. There is really no expectation that a molded case breaker with a magnetic trip element won't start to do something in that critical first 1/2 cycle. At that point all bets are off.

I don't have ready access to the IEEE Blue Book at the moment, but there is a discussion in it that became the basis of the 2005 NEC change. A I remember reading the material in question, it seemed to be more cautions than guidance.

Why chance it, if the existing installation suddenly finds itself subjected to more fault current than it is capable of interrupting/withstanding, either upgrade the system or reduce the available fault current back below the original equipment ratings. A 1:1 transformer can do wonders in limiting fault current.

 

[alehman]

Thanks David. I understand what you are saying, but my concern is that the new code provision may be misapplied by well-meaning engineers who are not knowledgable of the subtleties.

I'll dig out my Blue Book when I return to the office.

 

[stevenal]

But don't current limiting fuses limit the current even during the first half cycle? Even prior to fuse clearing, limiting action takes place. Worse case is that the breaker opens first within it's interrupting capacity.

 

[jghrist]

I'm not familiar with low voltage current-limiting fuses, but with medium voltage current-limiting fuses, manufactures publish maximum let-through current curves. For example, a 40A Cooper NX fuse will have a maximum let-through current of 5200A (peak) if the available rms symmetrical current available is 10000A. If the maximum let-through current is below the downstream breaker interrupting capacity, there should not be a problem.

 

[dpc]

jghrist,

The problem is that these "let-through" curves are just not reliable when you start looking at real-world situations.

Similar curves are published for low voltage fuses. But when the fuses are tested in series with a circuit breaker, the results don't always agree with these curves. The problems occur when the fault current isn't high enough to allow the fuse to operate in its current-limiting range. Sometimes the downstream breaker has enough impedance to limit SC current to a value where the fuse is no longer current-limiting and the breaker is subjected to first 1/2 cycle currents that exceed its rating, even though everything looks fine on the let-through curve.

In the old days, we called this the "up-over-down" method. But it was dis-credited back in the 1980s. That's why UL and NEC require that the fuse and breaker be a "tested combination".

But in reality, if you look at the breaker manufacturer's listing of tested combinations, most US-made current-limiting fuses seem to be on the list.

The problems have occurred when trying to retrofit a fuse upstream of some 50-year old molded-case circuit breaker made by I-T-E or some other defunct company. Maybe this is what the 2005 NEC revision is trying to address, but as a consultant, I'd be cautious about taking on that responsibility.

 

[mc5w]

Also, quite a few older systems are ungrounded 3-wire 3-phase or corner grounded 3-phase which poses a few extra problems besides interrupting rating of the circuit breaker.

I would NOT use an untested combination for over 600 volts because the definition of current limiting is quite different at over 600 volts. Under 600 volts current limiting for fuses is about a 1/4 cycle clear time. For over 600 volts a 2 cycle clear time is "current limiting" compared to a 4 cycle clear time.

Series combinations under 600 volts and less than or equal to 200 amps for class J fuses are well understood and test for the vast majority of circuit breakers that are still being made. If say you have an obsolete SquareD or manufacturer you can usually have their PEs help you with a combination that works.

For series combinations over 200 amps at under 600 volts things are a bit more iffy particularly for class L fuses. For class L fuses the downstream circuit breaker needs a lot more than a 10,000 amp or 18,000 amp interrupting.

There is also an NEC restriction as to how many motors you can have connected downstream of a current limiting fuse. Motor contribution to short circuits has to also be considered.

 

[tulum]

Thanks for the answers everyone.

Regards,
TULUM