NEC 2020 705.12(B)(1) Feeder VS (2) Tap

Status
Not open for further replies.

Xamacho

Member
Location
Texas
Occupation
Electrical Engineer, PE
I am afraid I am making some clients spend more money than need to be and I could use clarification. Right off the bat - Question #1) Is the Tap connection a smarter or cheaper solution than the Feeder connection, in general?

My situation in which I think I am adding more equipment than necessary.

To the left, a 600 amp switch in a Service panel. To the right, (open door) 600 amp MLO. In the metal wireway, three #4/0 per phase. Due to constraints, we cannot have a supply side interconnect or a loadside interconnect into a busbar. The options are either a 705.12(B)(1)"Feeder" or a (B)(2) "Tap" Connection and it will occur in this metal wireway (~10% filled). The PV is 340 amps at 125%.

20201009_141831cropped.jpg

A search in all my Mike Holt PV books shows an image like the one below (from 2014 code, but same concept) and this is the Option (b) using OCPD instead of upgrading the Feeder Conductors.

changes-to-2014-nec-interconnection-rule-70512-8-638.jpg

Due to this past history with this Mike Holt inspired image, I thought, "let's add a new 600 amp OCPD and PV interconnect (with the Feeder conductors) on the Line Side Lugs of this new 600 OCPD. Then on the load side, the feeder conductors continue on to the existing MLO panel.

Question #2) Is this the most expensive design possible and should we simply install a 600 amp main into the MLO Square D panel and interconnect into the Line Side lugs?

I always interpreted "other than the opposite end of the feeder" to mean "Must Feeder Connect and not Tap Connect", but I now realize this a misinterpretation? This language is only in the (B)(2) Feeder section, and by definition, at least half the taps are in the middle of the feeder. Non?

To only mildly complicate this, it's a 208/120 service and the inverters are 480, thus we have a 208 Delta Primary (Utility) : 480/277 Wye Secondary (PV) Transformer right after the PV AC Disconnect. I understand the 10', 25' rules surrounding Taps and even 240.21(B)(3) Taps Supplying a Transformer, but perhaps I thought it wasn't the appropriate solution. I know designing using "Feeder" connections is sound, I just don't like the idea waisting clients money when I don't have to.

Thanks for analyzing this situation and I apologize if it's already been addressed in previous posts.
 

jaggedben

Senior Member
Location
Northern California
Occupation
Solar and Energy Storage Installer
Where to start...

First of all, 705.12(B)(1) and (2) are not different options. They are two requirements that both must be met if you are tapping a feeder. B(1) pertains to the feeder, and (B)(2) pertains to any taps on the feeder, whether they are power sources or loads. (At least that's how I interpret (B)(2). It's terribly worded and I submitted a PI to try to get it fixed in the code and they slightly changed it for the better in the 2020 but not much.)

So, if you tap the feeder for the PV, and you insert a 600A fused disco downstream of your tap to protect the downstream feeder and panel, you have complied with (B)(1). You cannot tap the feeder without the OCPD downstream because you could overload the downstream feeder and panelboard.

Then there's the tap for the PV. Let's say it's a 25ft tap, it needs to be sized for not less than on third of 600A+340A. That number is still just below your 340A required for your output, so you wouldn't have a problem there, assuming you can achieve a 25ft tap. So that takes care of (B)(2). Your tap goes to a fused disco for the PV. Beyond that disco none of the rules discussed so far are in play anymore. (So your transformer is a non-issue with respect to the above.)

BTW if you do it this way you need to tap all three conductors per phase.

As far as whether you wasting anyone's money, I only see one option discussed so far, which is a feeder tap. So you'd need another option to compare to.

Another option might be to insert, say, a 1000A panelboard in between your main and sub. You'd put two breakers in it, one for the existing panel and one for the solar. You'll just have to price the components to see which set of them are more expensive, as well confirming if other costs (wire, conduit) are all equal depending on how far you have to go to find space for those equipment options. (I don't really see such space in the photo.)
 

Carultch

Senior Member
Location
Massachusetts
I've generalized your diagram, and will be using the labels I added:
1617241195536.png


Prior to adding the PV, Breaker A is the only OCPD that protects the MLO subpanel. Panels are routinely populated with more than the busbar rating worth of branch circuits, with the underlying reason you can do this, being that (1) not all of them will draw full load at once, and (2) the breaker protecting the panel will trip before overload among the branch breakers overheats the panelboard.

When you introduce the PV system, breaker C will equal 125% of X, and then be rounded up to the next standard size. In terms of considering whether the feeders are protected, it is the breaker size that matters, rather than what X is, or what 125% of X is. So even though there is a margin of safety on how C relates to X, we assume that current I2 will be as large as C, for determining if a circuit is protected. The section of conductor for the line side of C, is what is considered a tap. It is not yet protected at its ampacity. It is protected in excess of its ampacity by Breaker A. Breaker A protects it against short circuit current, while Breaker C protects it against overload.

Where currents I1 and I3 share the same circuit, the currents do not combine, due to being in opposite directions. This section of conductor just needs to be rated to carry the larger of I1 or I3. This will be the case by default, if it is already sized according to Breaker A.

Without adding Breaker B, the current I4 equals I1+I2. If the subpanel draws enough load, I4 could equal A+C. This is not acceptable, unless the feeder and subpanel busbar have enough ampacity for A+C. So you can either upgrade the feeder and replace the subpanel, or you can insert Breaker B between Breaker C and the subpanel. It doesn't necessarily need to be a breaker, just an OCPD in general, so a fused disconnect could also work. It could be the main breaker of the subpanel, or it could be a separate device, and meet the intent of this rule.

If everything is new, I'd recommend internalizing Breaker B as the main breaker of the subpanel, so it is one less piece of equipment to install. In an existing situation, Breaker B would most likely be added as a separate device. Because adding a main breaker inside an existing MLO panel is usually not practical.
 

wwhitney

Senior Member
Location
Berkeley, CA
Occupation
Retired
Then there's the tap for the PV. Let's say it's a 25ft tap, it needs to be sized for not less than on third of 600A+340A.
While subsection (2) can be read to apply to power source output connection itself, that makes zero sense from a physics point view. So I would say section (2) only applies to taps that can be supplied by both the utility source and the interconnected power source. Section (1) adequately covers the interconnection of the power source itself.

Cheers, Wayne
 

jaggedben

Senior Member
Location
Northern California
Occupation
Solar and Energy Storage Installer
While subsection (2) can be read to apply to power source output connection itself, that makes zero sense from a physics point view. So I would say section (2) only applies to taps that can be supplied by both the utility source and the interconnected power source. Section (1) adequately covers the interconnection of the power source itself.

Cheers, Wayne

I don't follow. My interpretation is the only way I could make sense of it 'from a physics point of view.'

As best I can tell, the only reason for taps to be sized per 204.21(B) in the first place is if they short to each other or ground. Oversizing provides less resistance, making the upstream oversized OCPD more likely to operate at least at its short time rating. Perhaps there's also some additional mechanical robustness against damage from the short. It has nothing to do with operating current which is limited by the overcurrent device at the end of the tap.

If multiple sources can feed the feeder and tap, then they can feed a short circuit on it regardless of whether the tap was made for a power source or a load. In other words, any tap made on such a feeder "can be supplied by both the utility source and the interconnected power source."
 

wwhitney

Senior Member
Location
Berkeley, CA
Occupation
Retired
Ah, you are right, the tap rule minimum sizes are for protection against short circuit and ground fault, and so what matters is the available current sources during a fault, not what current the tap can carry during operating conditions. And during a fault, the power source output conductors can carry current from both the utility and the power source.

I was only thinking about what they could carry during normal operating conditions. Thank you for clarifying that.

On related note, the requirement of section (2) seems off. I assume the tap sizing rules have to do with conductors surviving the available fault current (AFC, not sure if that's the usual acronym). And the AFC from the utility is a large multiple of the utility side OCPD rating, while the AFC contribution from the power source, when it's an inverter, would be about the same as the OCPD rating, yes? So it hardly makes sense to sum the two equally, it should be more like the utility side OCPD plus 1% of the inverter output rating.

Cheers, Wayne
 

jaggedben

Senior Member
Location
Northern California
Occupation
Solar and Energy Storage Installer
....

On related note, the requirement of section (2) seems off. I assume the tap sizing rules have to do with conductors surviving the available fault current (AFC, not sure if that's the usual acronym). And the AFC from the utility is a large multiple of the utility side OCPD rating, while the AFC contribution from the power source, when it's an inverter, would be about the same as the OCPD rating, yes? So it hardly makes sense to sum the two equally, it should be more like the utility side OCPD plus 1% of the inverter output rating.

Cheers, Wayne

I agree, at least in principle. Of course, some sources, such as battery inverters, may be able to deliver more fault current than others, such as solar inverters. But my real criticism of your idea is that would be way more scientific than the 240.21 tap rules are in the first place. And we can't have that. ;)
 
Status
Not open for further replies.
Top