Fault Current Limiting Breakers and downstream disconnection times.

[healyx]

I know that if a circuit breaker is interrupting a known symetrical fault, you can find out the disconnection time by looking at the breaker's time-current curve (Given you have calcualted the prospective short circuit current (PSCC) using cable impedance etc.). Can you still calc disconnect time this way if a fault current limiting breaker is placed upstream of the breaker closest to the fault? The upstream breaker will/may reduce the peak current so that the downstream breaker will take longer to disconnect. Anyone know how to work this out? The peak limited current is quoted as Asymmetric (I think) where as the PSCC is a symetrical RMS value used to index the time-current curve for the breaker. I wouldn't have thought you could use the asymetric limited peak current with the breaker time-current curve to determine the "new" disconnection time??? I have noticed that for lower fault values the limited peak kA is higer than the PSCC kA value (obviously due to RMS calc). Anyone know how to find the actual disconnection time? (FYI, I am trying to work out if an upstream FCL breaker could cause a downstream breaker to move into the thermal (slow) region instead of magnetic (fast) region due to the FCL action and hence cause an earth fault to not be cleared fast enough).

Cheers.

 

[rbulsara]

Two breakers in series will see the same current at any given time, so it is only a matter of comparing independent time curves at a given current to see which one will operate faster.

If the upstream breaker "limits" the peak, it means that it already has operated (opened up). There is nothing for the downstream breaker to do after that.

I think you are mixig up concept of series rating for short circuit current rating analysis with time coordination, which by itself is another debate.

Rafiq Bulsara

http://www.srengineersct.com

 

[healyx]

Thanks for the reply Rafiq.

I know you can definitely use the time-current curves for breakers in series when the upstream breaker is NOT current limiting type.
I was of the view that you weren't supposed to use time-current curves only for selectivity analysis with FCL breakers because the limiting effect of upstream breakers changed the dynamics of the downstream breakers time-current behaviour. Better to use manufacturers table for this.

Therefore I assumed disconnect time on the downstream breaker would be altered in a similar way.

BTW, as an aside, if the upstream breaker "limits" the current it doesn't follow that the downstream breaker has nothing left to do. I thought it was common for the upstream to limit let-through but the downstream still is the only one that opens (upstream breaker just reduces downstream breakers exposure to full fault current).

I understand the concept of short circuit current rating analysis, I just want to know if an upstream FCL breaker will "shift" the time-current curve of a downstream breaker due to limited fault current.

It seems logical to me that if the upstream breaker is blocking fault energy, then the lower breaker won't operate as quickly - but maybe the timing is critical. Maybe the downstream breaker sees enough of the initial fault to allow the use of calculated PSCC to determine disconnection time and energy limiting applies after that??? Can anyone explain this to me?

 

[davidbeach]

What are you really trying to accomplish?

If you are leaving a reasonable amount of coordination interval between breakers any change in time due to the presence of a current limiting breaker shouldn't impact the coordination.

The only way a current limiting circuit breaker can have any current limiting effect is if there is contact parting. Some current limiting breakers are known to "burp" in that there is contact movement but not the trip is never released.

If it is critical to know exactly what is happening then you probably need to have the combination tested at the prospective fault current of interest. If it were possible to calculate what is going to happen, the circuit breaker manufactures would be able to derive series rated combinations by design rather than having to rely on testing.

 

[dpc]

There is a misconception about how current-limiting breakers and fuses work. The only way a current-limiting breaker is considered to be current-limiting is because it interrupts the current before the first current peak occurs (within first 1/2 cycle). The classification as a current-limiting device does not imply that the device is doing anything to reduce the available fault current, it is simply operating so quickly the fault current never reaches a peak value. Current-limiting fuses work the same way.

So as Rafiq indicated, if the fault current is high enough, the breaker is going to trip and there is nothing to be done about coordination, because it is not possible to coordinate.

In fact, it is generally not possible to coordinate any two molded-case circuit breakers in series for high values of fault current because both breakers will trip on instantaneous element.

"Theory is when you know all and nothing works. Practice is when all works and nobody knows why. In this case we have put together theory and practice: nothing works... and nobody knows why! (Albert Einstein)

 

[healyx]

"What are you really trying to accomplish?"

I need to know disconnection time for fault loop impedance calculation. Standards require that an earth fault clears within a certain time.

If for a given earth fault current, the breaker is marginally in the magnetic region it will trip in less than 100ms. If the earth fault current is limited cause the downstream into the thermal region it could take multiple seconds to disconnect - which would violate required disconnection time limit.

I just wanted to know if this was possible - i.e. I do a design and verify earth fault will disconnect in under 0.4 seconds or 5 seconds - only to find later that the influence of an upstream FCL breaker pushed it into thermal region and earth fault takes 8 seconds to clear. Note - I haven't found a case where this would be marginal yet, usually the fault currents are too low for limiting to kick in in the upstream breaker - but given peak current is asymetrical and PSCC is symetrical I don't know when limiting actually kicks in (e.g. A circuit breaker I am looking at has a PSCC to Peak current curve that starts at 5kA PSCC and limited peak is listed as 6.5kA. It has no data below 5kA PSCC so does that mean no limiting beneath this point? and would you say 6.5kA peak for a 5kA PSCC has limited at all?)

 

[rbulsara]

healyx:

Asking the same question in different words, would not yield a different response. You have a great misconception about how "current limiting" devices work or what does the term mean. They are not current "reducing" devices.

Rafiq Bulsara

http://www.srengineersct.com

 

[healyx]

Thanks to all for your responses.

It think that clears it up. FCL breakers only "appear" to limit current because they trip before top of the waveform.
So this has raised another question, how does cascading work? E.g. Data sheets show I can use a 36kA 250A FCL breaker to protect a downstream 6kA 63A breaker upto 30kA with discrimination achieved up to 30kA (i.e. if 30kA fault only the 6kA 63A breaker will trip). In this case does the upstream breaker do anything? If it does nothing why can't I expose the 6kA 63A breaker to a 30kA fault without upstream FCL breaker? Is the upstream breaker pre-arcing or something? Therefore limiting let-through energy but not altering the current that the downstream breaker sees?

 

[davidbeach]

You're obviously in IEC land. I have no idea how "cascading" works, but if it is anything like series ratings, which allow a higher rated device and lower rated device in series to be used where the available fault current is above the rating of the lower rated device, you have to anticipate that both devices will open. The only mechanism for current limiting is opening. The "burping" effect I mentioned earlier happens at currents below the range of current limiting. If the breaker is going to current limit it is going to open. No way that a 6kA breaker can interrupt a 30kA fault without the upstream current limiting breaker opening. No other means of current limiting available.

 

[rbulsara]

I believe cascading is same as the series rating in the USA. The series rating combination in the USA is only approved by testing by a NRTL such as UL. There is no mathematical calculation or analytical method for it. Manufacturers have a published list of series rated combination of devices and only those can be applied.

The testing only ensures that for the rated short circuit current, either or both devices will open without themselves failing dangerously. There is no guarantee of coordination nor one is expected.





Rafiq Bulsara

http://www.srengineersct.com

 

[healyx]

Just to be clear I am have only been discussing low-voltage IEC compliant devices, not NEMA breakers. Perhaps that has caused some confusion.

Here in Australia (our standards are based on IEC standards) breaker manufacturers publish cascading and selectivity (discrimination) tables and state that their FCL breakers have "enhanced selectivity". This is where a downstream breaker can be placed in a higher fault level if protected by an upstream FCL and it guarantees that the lower breaker will trip first even when exposed to a fault higher than it's native level e.g as I stated before a 6kA breaker can withstand a 30kA fault if a 36kA breaker is upstream and the manufacturer states that the 6kA will be the only breaker to latch open if the fault level is lower than 30kA. To do this the upstream breaker does partially open - to quote Schneider Electric (Merlin Gerin):

"due to the short-circuit current (electrodynamic forces), the contacts in both devices simultaneously separate. The result is major limitation of the short-circuit current;
The dissapated energy provides the reflex tripping of the downstream device, but is insufficient to trip the upstream device"

Hence back to my original question: If the upstream breaker is partially opening does it significantly alter the disconnection time of downstream breaker. I think I have satisfied myself that if the upstream breaker is even partially open then the downstream will still disconnect very quickly (definitely in the magnetic region).

 

[dpc]

If the contacts part in the upstream breaker but the breaker does not trip, the resulting arc will introduce resistance into the circuit reducing the current. I would think this would cause the downstream breaker to possibly operate more slowly, certainly not faster.




"Theory is when you know all and nothing works. Practice is when all works and nobody knows why. In this case we have put together theory and practice: nothing works... and nobody knows why! (Albert Einstein)

 

[healyx]

Thanks dpc - I meant to say in my last post - if upstream breaker is parting then the fault must still be very significant to the downstream breaker and certainly significant enough to make it trip in the magnetic region - but no, as you have suggested it would be a bit slower due to arcing resistance upstream.
Gets back to my original question - which is becoming theoretical - if an upstream breaker is dissapating energy by arcing, is the downstream breaker's time-current curve (using prospective short circuit current at fault point as the current index) still valid to determine disconnection time? Is there any way it could be slowed to the point whereby it trips in the thermal region rather than the magnetic region (if near the thermal/magnetic threshold this could be the difference between 0.1s disconnect time and 8s.)

 

[dpc]

In most situations, at least with UL (US) type MCCBs, if there is enough current to blow apart the contacts on the upstream breaker, the smaller downstream breaker would nearly always trip in the instantaneous region. By the time the breakerr contacts parts, the smaller breaker would already be in the tripping sequence, although it may not have fully tripped. This would depend on the size difference between the two breakers. If you are talking about a 600 A and a 400 A breaker in series, I still think both will trip 95% of the time for a high current fault. If it is a 30 A breaker in series with a 600 A breaker, there is a better chance of coordination.

"Theory is when you know all and nothing works. Practice is when all works and nobody knows why. In this case we have put together theory and practice: nothing works... and nobody knows why! (Albert Einstein)