Minimum conductor length to keep fault current under 10k with #8

[JimmysLimeade]

I recently got this rejection from APS:
Not really sure how to get this number, 110.10 doesn't really help with this and I can't find anything online. Are there any APS proficient people that can help me figure out how to get the correct length? Thanks!

 

[wwhitney]

There's more details and subtleties to available fault current calculations than I'm familiar with, but here's a basic answer to start with:

Say you know that at a certain point in the electrical system, if you short two wires with a nominal 240V potential between them, you'll get a fault current of 21,188A. [APS's response implies this is true at the terminals of the PV breaker in the service panel, and I question whether that's a sharp number or just a conservative estimate; if the real value is lower, that would make things easier for you.]

Then Ohm's law tells you that the source impedance at that point is just R = V/I = 240V / 21,188A = 11.3 milliohms. To get the available fault current down to 10,000A, you'd need a source impedance of R = V/I = 240V / 10,000A = 24 milliohms. So to do that you need to add 12.7 milliohms of impedance.

If you want to do that just relying on the #8 Cu conductors between the service panel and the unfused disconnect, then per Chapter 9 Table 9 in the NEC #8 Cu has an impedance of 0.69 milliohms/ft (for 3 wires in PVC conduit at a 0.85 PF; whether those conditions are reasonable to use are one of the subtleties I'm unclear on ). That means to add 13 milliohms, you'd need a roundtrip conductor length of 12.7/0.69 = 18.4 ft.

So this simple calculation says that you just need to move your PV disconnect to be 9.2 ft or more of one-way conductor run from the service panel, rather than only 6ft. However, I'm unclear on whether considering a 120V line to ground would be the controlling case with a different answer, or if there's something else I'm overlooking. So hopefully someone more experienced will comment.

Cheers, Wayne

 

[don_resqcapt19]

Download Bussmann's "FC²" app.
It looks like they are saying the available fault current at the supply end of the 8 AWG conductors is 21,188 amps.
That app shows 10' of 8AWG copper in a ferrous raceway will result in 9,929 amps of fault current at the load end of the conductor.

 

[hmspe]

SRP (Salt River Project) has different fault current values than Arizona Public Service (APS). If this really is in an APS area the correct fault current value to use at the service is 14,318A symmetrical. If this is SRP then 21,188 is correct and you need to take references to APS off the drawings.

 

[wwhitney]

SRP (Salt River Project) has different fault current values than Arizona Public Service (APS). If this really is in an APS area the correct fault current value to use at the service is 14,318A symmetrical. If this is SRP then 21,188 is correct and you need to take references to APS off the drawings.

So those power companies specify that the installer is to assume those AFC values for any residential service, based on their typical primary impedance, transformer impedance, and service conductor length? As say, the 99% worst-case AFC, a value that very few residential services would exceed in reality?

At what point is the specified AFC to be assumed, at the meter socket? And then when doing AFC calculations, is any impedance assigned to the service OCPD and panelboard, or are only conductor lengths considered?

Thanks,
Wayne

 

[augie47]

It would be interesting to know the details leading to the 21,288 at the meter (service conductor, length,etc).
That would be an usually high number in this area for a 200 amp residential service.

 

[letgomywago]

It would be interesting to know the details leading to the 21,288 at the meter (service conductor, length,etc).
That would be an usually high number in this area for a 200 amp residential service.

Need to be a big transformer and short underground to get that at the meter.

 

[jaggedben]

(First of all, APS is totally full of you-know-what having any input into conductors and disconnects on the load side of a 200A resi service, just because there's solar. Do they do this check for air conditioners? But there's probably no effective way to tell them to ---- off, so...)

They said it's insufficient for nonfused disconnects rated 10k. So change it to a fused disconnect, send it back, and see what they say.

Don't waste your time on calculations. Or at least, not unless they send it back again.

 

[electrofelon]

And just a reminder guys there is no requirement in the NEC to use the fault current value provided by the utility. I have climbed poles and pumped out UG vaults to get the actual transformer data.

 

[ron]

They are implying that there is 21,188A of available fault current at the main service panel. With that in mind, there better be 22kAIC rated breakers in there.
As several have said, you can add conductor length or add fuses to the disconnect to raise its SCCR. In reality, I would bet a standard 2 pole safety switch is only rated at 5kA, but I don't want to make more trouble.

 

[hmspe]

The local requirement in the Phoenix area is that the AFC provided by the utility is always used. Municipal plan review will reject fault current calculations based on transformer impedance and service conductor length. The table AFC value is at "the service".

I don't think I've ever seen a residence that would come close to the SRP values. A 400 amp servise with CTs is less that a class 320, at 18.976 amps.

My office was in the Phoenix area for 25+ years. Moved to Texas to "retire" 5+ years ago. The calls haven't stopped and almost all my work is in Arizona.

 

[electrofelon]

They are implying that there is 21,188A of available fault current at the main service panel. With that in mind, there better be 22kAIC rated breakers in there.

There is almost certainly a series rating between the main and the branches and also almost certainly the main is rated at least 22k - but is possible this is old or oddball.

As several have said, you can add conductor length or add fuses to the disconnect to raise its SCCR. In reality, I would bet a standard 2 pole safety switch is only rated at 5kA, but I don't want to make more trouble.

I believe all NF switches are series rated 10k with any breaker. They may have higher ratings depending on the breaker. For example here is Joseph Eaton's chart:

https://www.eaton.com/content/dam/eaton/products/low-voltage-power-distribution-controls-systems/swiching-devices/canada/eaton-combination-short-circuit-current-ratings-for-non-fusible-safety-switches-technical-data-td00801005e-ca.pdf