Fault current reduction for lighting control

[davidr43229]

My reading of the brochure is that "let-through" is not acceptable, at least with Square D equipment. Keep in mind that this is an early document, it's obviously from the marketing department, and Square D makes breakers, not fuses.

Let-through is very acceptable, that is how you protect your downstream components.
Just my $.02

 

[jim dungar]

Let-through is very acceptable, that is how you protect your downstream components.

Let-through can only be used to protect equipment that does not try to close onto or open a fault such as bus bars, conductors, and maybe isolation switches. There are very few (if any) breakers and fuses <100A that will not begin to open at fault levels as low as 5kA.

 

[ron]

You may only use let through from a current limiting device to determine a SCCR if it is series rated. It can only be series rated if it is a tested combination by a NRTL.

 

[winnie]

Let-through can only be used to protect equipment that does not try to close onto or open a fault such as bus bars, conductors, and maybe isolation switches.

What about things such as time clocks and receptacles, as per the original post? These are things which have separable contacts, but which are not expected to open the circuit automatically during a fault.

-Jon

 

[davidr43229]

What about things such as time clocks and receptacles,

Any passive component can be protected, which all the above listed are.
Even a Class CC will go into it's current limiting region at 5,000amps.
http://www.cooperbussmann.com/pdf/1023.pdf
Just my $.02

 

[hmspe]

Any passive component can be protected ...

I'm not trying to be argumentative, but based on what? I've seen a number of opinions posted, but "David said in Mike Holt's forum that it's OK" won't get me very far as a response to a plans reviewer or an inspector. I'm not disagreeing with your statement, just wanting to find unbiased documentation (meaning written without a marketing agenda) to back up the opinions. The fuse sheet referenced lists let-through values, but says nothing about what the let-through values can be applied to. "Passive component" in the 2005 NEC shows up only in Article 504, which is Intrinsically Safe Systems. How do we determine what's "passive"? If there's a breaker (meaning dynamic impedance) in a circuit is let-through still applicable to a downstream passive component? Whether your answer is yes or no, the follow-up question is, "Based on what?"

Martin

 

[davidr43229]

<deleted, duplicate>

 

[davidr43229]

I would encourage you to read this entire article from EC & M
Listing a device to UL 508 often requires additional protection for the product to pass the test sequence. The overcurrent device, branch circuit breaker, or fuse is then required to be marked on the Listed UL 508 device. As noted earlier, the device must have a minimum of 1,000A SCCR to meet UL 508. Some relays have a 5000A SCCR when used with and RK5 fuse. In order to ensure a NEC-compliant installation, the contactor must be installed per the manufacturer?s instructions with the required circuit breaker or fuse for protection.
http://ecmweb.com/news/electric_know_really_protect/index.html
or :
http://www.comnews.com/stories/articles/1006/1006circuit_breakers.htm

According to the NEC 110.10, overcurrent protection devices shall be selected to permit the OCPD to clear a fault without damage to the electrical components of the circuit. By reducing the amount of energy that passes through to the protected device, the damage that reduces repair and downtime is reduced. In order to successfully protect sensitive equipment, the upstream overcurrent protective device needs to be able to operate in a short amount of time, and consistently limit the amount of fault current/energy that passes through to the downstream devices
You can do a google or yahoo search for current limitation, protecting downstream components and read more articles.
Just my $.02

 

[Pierre C Belarge]

What about things such as time clocks and receptacles, as per the original post? These are things which have separable contacts, but which are not expected to open the circuit automatically during a fault.

-Jon

Receptacles and timeclocks are electrical equipment that is generally installed on a Branch Circuit. They are generally installed along the point of the System where SCCR is not considered as a possible issue.

I can see how a time clock or receptacle installed in close proximity to a panel that has higher than 10,000 AIR could become an issue (especially with the transformers installed closer or within a building these days). I have been discussing this with electricians for the past year, mostly to blank stares and no response. The circumstances that I come across in my jurisdiction is that it would seem as though no one is really paying any attention to this issue.

So, this topic is of great interest to me. I also have been reading on this topic, but I am not an expert. I do want to understand more...hopefully we will see more info posted here about this.

 

[hmspe]

http://ecmweb.com/news/electric_know_really_protect/index.html
or :
http://www.comnews.com/stories/articles/1006/1006circuit_breakers.htm

Thanks for the links. I especially liked last line of the second article:

Kevin S. Arnold and Cooper Bussmann are with Server Technology, Reno, Nev.

Sorry. Engineer's humor... Of course, there probably really is someone somewhere named Cooper Bussmann...

At this point my take is that:
1. We can apply contactors and time clocks fully rated, based on fault calculations. That may require increasing conductor lengths to reduce fault current values at the device.
2. We can apply some devices at higher fault values if the manufacturer provides listings for specific upstream fuse types. We may still need to increase conductor lengths, but the acceptable fault current may be 5KA rather than 1KA.

One open question is whether adding a current limiting fuse ahead of a device allows using the let-through of the fuse as the basis for downstream fault calculations. In the EC&M article, in the paragraph starting "Overcurrent protective devices", it says, " The combination of the circuit and the overcurrent device must limit the available let-through current to a value that doesn't exceed the SCCR of each listed device." However, the article goes on to discuss UL 508 listings, and specifically discusses having listed combinations (fuse and device) with higher SCCR ratings. The higher rating is presented as a device rating, not a maximum fuse let-through value. If fuse let-through was the controlling parameter wouldn't they list something like 'maximum let-through for upstream RK5 fuses' rather than a device rating when used with RK5 fuses?

Martin

 

[hmspe]

I can see how a time clock or receptacle installed in close proximity to a panel...

How about the receptacle close nippled to the panel at most residences? 18" of #12 or #14 wire doesn't drop the fault current much. Where I live we have to assume 10KA availabel fault for a 200A service and 22KA for a 400A, and most services are "all in ones".

History (based on hundreds of thousands of installations) would seem to say that this practice really isn't all that dangerous, but it is a technical violation.

Martin

 

[davidr43229]

At this point my take is that:
1. We can apply contactors and time clocks fully rated, based on fault calculations. That may require increasing conductor lengths to reduce fault current values at the device.
2. We can apply some devices at higher fault values if the manufacturer provides listings for specific upstream fuse types. We may still need to increase conductor lengths, but the acceptable fault current may be 5KA rather than 1KA.

1. Absolutely yes, either increasing wire lengths, or introducing an Overcurrent Protective device that exibits current limitation, down to the SCCR levels that you require.
2. also correct

If fuse let-through was the controlling parameter wouldn't they list something like 'maximum let-through for upstream RK5 fuses' rather than a device rating when used with RK5 fuses?

UL has a list of Maximum levels, however most manufacturers have lower levels than the stated maximum. They are public record.
http://www.sacramentoelectronics.com/bussmann/pdfs/special_interest/SPD.pdf
Try page 59 to 64.

 

[Pierre C Belarge]

How about the receptacle close nippled to the panel at most residences? 18" of #12 or #14 wire doesn't drop the fault current much. Where I live we have to assume 10KA availabel fault for a 200A service and 22KA for a 400A, and most services are "all in ones".

History (based on hundreds of thousands of installations) would seem to say that this practice really isn't all that dangerous, but it is a technical violation.

Martin

For residential applications:
I will not disagree with your numbers, but I will say that in our area the Available fault current for a 200 amp standard service is in the range of 3000-6500 available short circuit current. I do not see that as a problem for receptacles installed via a nipple at the service panel...although I could easily be proved incorrect, as I have never performed the calculations to back up my statement.

For the larger services, I agree that in the newer construction, the transformers are generally closer to the building/service and the available short circuit current available is generally higher. Again I have not performed the calcs, so I am not sure how much higher.

 

[hmspe]

UL has a list of Maximum levels, however most manufacturers have lower levels than the stated maximum...

I use Bussmann's SPD regularly. Excellent resource. Unfortunately, this doesn't really answer what I was trying to ask. I'm probably not being clear enough in my questions, or maybe I'm leaving too much out trying to keep the posts short. I know about fault reduction based on conductor length and impedance. I know about current limiting fuses and where to find let-through values. I know that the SCCR listing for a contactor per UL 508 may be higher if a specific protective device is upstream.

Here's a sample condition that I can't document an answer for: Assume the available fault current at a panelboard is 42KA. On a 20A lighting circuit, can I insert a 30A RK5 fuse betwen the branch circuit breaker and the contactor, and base the fault reduction on the 2KA maximum let-through of the fuse instead of the 42KA fault current at the panel? (Let-through data is taken from page 60 of the 1996 Bussmann SPD for FRN-R fuses, and 30A is used because it's a table value, not an interpolated value.) What do I cite (NEC, UL, college text book, manufacturer's data, etc.) to justify the answer? How would I convince an inspector that this was (or was not) acceptable? As I previously posted I can find a lot of opinions and a lot of sweeping generalities, but very little that provides calculations, test data, code referebces, or standards references demonstrating that the approach is valid.

Just for fun I ran fault reduction calcs for the above using the method found in in the Bussmann SPD. Dropping from 42KA to 1KA requires 37' of #12 copper. Dropping from 2KA to 1KA requires 19' of #12 copper. Dropping from 42KA to 5KA (contactor SCCR rating in combination with an RK5 fuse) requires 7' of #12 copper.

I appreciate all the comments this thread is generating. Thanks for taking the time to post.

Martin

 

[davidr43229]

Try using a CLass CC fuse . At 42K available a 30 amp fuse will let through 1,700, then introduce your wire.....or a Class T (JJN-30), at 42K available the let through is 1,200amps, then introduce your wire.
http://www.bussmann.com/pdf/1026.pdf

or Try a DFJ , at 42K, will let through 800amps at a 30 amp fuse and a 20 amp fuse will let through approx 600amps (interpolated).
http://www.bussmann.com/pdf/15cc9ab7-72a6-4460-b136-28f4a7d71037.pdf
For you to go from 42Kamps down to less than 1,000amps takes a huge amount of current limitation, within a confined space.
You can use the Manufacturers data to show the AHJ.
All the fuses I have shown you have standard disconnects to hold them.
Just my $.02

 

[winnie]

I know that I am jumping in to Martin's thread with my own questions...but I think that I see where he is going.

I have a device with an SCCR rating. The SCCR rating is substantially less than the available short circuit current at a given panel.

I place something called a 'current limiting fuse' upstream of my 'device'. This current limiting fuse has a 'let through' current that is less than the 'short circuit' rating of the device. On first brush, it seems completely obvious that if the let through current of the fuse is less than the short circuit rating of the device, then we've provided a properly protected system.

But then someone brings up the requirement for series ratings of OCPD systems. Say you have a system with 75 kA of available short circuit current, and you have 10kA circuit breakers. You _cannot_ simply use a current limiting fuse upstream of the breakers and call things good; you must use a 'series rated' system that has actually been tested as working safely. Clearly, in at least some cases, the reduced 'let through' current of a current limiting fuse is _not_ considered sufficient to protect downstream devices.

My understanding of this situation is that because of dynamic impedance any downstream device which attempts to open the circuit during the fault might in essence 'see' far more than the 'let through' current of the fuse, and that this interaction might cause the combination of current limiting fuse and downstream OCPD to fail. But if my downstream 'device' is not an OCPD, and is not attempting to open against the fault current, then the 'let through' current is a reasonable indication of the stresses that the device has to be able to tolerate.

However, having been burnt once with a misunderstanding of how to use 'let through' current (even though it seems pretty obvious to me), the question is: Where is the 'official' document that tells me that I can use the 'let through' current of a current limiting fuse as the available fault current for selecting proper SCCR ratings of downstream devices?

-Jon

 

[davidr43229]

But then someone brings up the requirement for series ratings of OCPD systems

Series rating has nothing to with this example. UL Series rating is used to protect existing CB that the available fault current has risen beyond the Withstand of that in stalled breaker.

My understanding of this situation is that because of dynamic impedance any downstream device which attempts to open the circuit during the fault might in essence 'see' far more than the 'let through' current of the fuse

Dynamic impedence is only a concern when the device opens within the 1st 1/2 cycle (.008) seconds.

Where is the 'official' document that tells me that I can use the 'let through' current of a current limiting fuse as the available fault current for selecting proper SCCR ratings of downstream devices?

The proper document is a Bussmann SPD. You asked.....
http://www.bussmann.com/library/docs/spd02/SPDSection10.pdf
Just my $.02

 

[winnie]

Series rating has nothing to with this example. UL Series rating is used to protect existing CB that the available fault current has risen beyond the Withstand of that in stalled breaker.

IMHO 'Series Rating' is a parallel situation, that is the only reason that I brought it up. There are techniques for calculating 'let through' current of a fuse; I wanted to confirm that these calculations could be used for time clocks, even though they cannot be used for circuit breakers.

The proper document is a Bussmann SPD. You asked.....
http://www.bussmann.com/library/docs/spd02/SPDSection10.pdf
Just my $.02

Yes, that does appear to be the proper document, thanks!

This initial parts of the document describe using 'up-over-down' graphs and current limiting fuses to protect things such as switches, receptacles, _conductors_, etc. There are also discussion of techniques for calculating the SCCR values suitable for conductors. Finally there is a discussion of series ratings. The only thing missing IMHO is a statement in the series rating section (starting page 11) saying why the techniques described on the previous pages are not applicable

-Jon

 

[hmspe]

Jon, thanks for jumping in. You did a good job of summing things up. And David, thanks for all your posts.

The 2002 SPD (which is linked in post #37) has a lot more information that the 1996 SPD (which was linked in post #32).

Martin

 

[davidr43229]

You're very welcome.
Dave