"Service Disconnect" wiring question

[Steve16]

For sake of argument lets say this is on the 2023 NEC where wording states:

(D) Service Disconnecting Means.

A disconnecting means in accordance with Parts VI through VII of Article 230 shall be provided to disconnect all ungrounded conductors of a power production source from the conductors of other systems.

Looking at plans for a dwelling with 41 mods with microinverters @ 1.6 each. 41 x 1.6 A x 1.25 = 82 A
Plans show #4 thhn with 90 fuses in a 100 amp disconnect
230.42(B) calls for the minimum ampacity of the conductors to be not less than the rating of the service disconnecting means
310.12 says for services and feeders you can size the conductors based at 83%
So is a #4 feeding this disconnect correct?

What if I have 50 micros @ 1.6A? Now were at 100A fuses and #3 thhn conductors
Can you still feed the disconnect using #4s with the 83%?

 

[wwhitney]

The 83% rule (310.12) would never apply to a service connected only to PV, as such a service is not supplying the "entire load associated with a dwelling unit."

As to 230.42(B), which references 230.79, again the service disconnect for the PV is not "the" service disconnect for a one family dwelling. [If 230.79(C) said "any service disconnect mounted on a one family dwelling needs to be 100A at least," then you'd have an issue.]

As to #4 THWN with 90A fuses in a 100A disconnect, that's fine for a 82A calculated "load" as long as all the terminations are 75C rated as expected. The 75C ampacity of #4 Cu is 85A, which is large enough, and 240.4(B) allows protecting those conductors at 90A.

Cheers, Wayne

 

[Steve16]

Thanks Wayne, that was my thought. Just trying to see it from all angles before kicking the plan review back.

Would the #3 cu be code compliant tapping the #4 cu service conductors from the meter to main panel? Or do they need to be upgraded as well? Seems like they would need to be minimum #3 as well if the potential is 100 amps feeding through the taps

 

[wwhitney]

Would the #3 cu be code compliant tapping the #4 cu service conductors from the meter to main panel? Or do they need to be upgraded as well? Seems like they would need to be minimum #3 as well if the potential is 100 amps feeding through the taps

Sorry, are you saying the existing service is a 100A residential service with #4 Cu conductors, using the 83% rule? They can be spliced with #4 Cu conductors for the PV, no need to upsize the existing.

[The term "tap" is often used to mean feeder taps under 240.21(B); that meaning would never apply to service conductors. There are no tap rules for service conductors, you mostly just size them to the load.]

Cheers, Wayne

 

[Steve16]

Sorry, are you saying the existing service is a 100A residential service with #4 Cu conductors, using the 83% rule? They can be spliced with #4 Cu conductors for the PV, no need to upsize the existing.

[The term "tap" is often used to mean feeder taps under 240.21(B); that meaning would never apply to service conductors. There are no tap rules for service conductors, you mostly just size them to the load.]

Cheers, Wayne

Yes, by tap I mean a splice. It's typically referred to as a line tap or line side tap of the service entrance conductors.

And yes, the service is 100 amp #4s using the 83% rule. But can the #3s connecting the solar 100 amp "service disconnect" splice into these #4 service entrance conductors by code?

 

[wwhitney]

And yes, the service is 100 amp #4s using the 83% rule. But can the #3s connecting the solar 100 amp "service disconnect" splice into these #4 service entrance conductors by code?

Yes, sure, because we established that you only need #4s for the solar, not #3s. So use #4s for the solar.

Cheers, Wayne

 

[Steve16]

Yes, sure, because we established that you only need #4s for the solar, not #3s. So use #4s for the solar.

Cheers, Wayne

The 82 amp continous load aside, if were at higher than 90 amps continous and sized at #3s fused at 100 amps, is that allowed to splice into the #4 service conductors or would they need to be sized up?

Edit: sorry 90 amps

 

[wwhitney]

The 82 amp continous load aside, if were at higher than 90 amps continous and sized at #3s fused at 100 amps, is that allowed to splice into the #4 service conductors or would they need to be sized up?

The 82A figure isn't a continuous load, the continuous load is just 41 * 1.6 = 66A. Then with the 125% continuous factor you (generally) need a conductor with ampacity 82A. So by 90A, presumably you mean 72A continuous, which becomes 90A ampacity after the 125% continuous use factor.

As to the question, I'd say it's ambiguous.

On the less conservative side, there's an argument that with a 100A residential service the provisions of 310.12 would still let you use #4s for the solar in the above configuration and protect them at 100A. [That argument got weaker in 2023 with treating the PV line side connection more like a service; it depends on treating it more like an adjunct to the residential service, and contradicts the first sentence of my first post in this thread. But I could still make that argument.] In practice that would be fine, as the actual continuous load would be under 80A, and the #4s themselves are good for 85A continuous at 75C.

If we reject the 310.12 argument, then the answer is that any service entrance conductors (SECs) supplying the PV service would need to be #3s, but existing service conductors that are not SECs can stay #4s. That's because SECs do have a 125% continuous use factor for ampacity, while other service conductors don't. This is per Article 230, and I'm a bit hazy on the exact demarcation point between SECs and service conductors that aren't SECs.

Cheers, Wayne

 

[wwhitney]

But I could still make that argument.

Rather than make that argument, I'd prefer just to say that if you have a 100A residential service with #4 Cu service conductors, and you want to interconnect 80A continuous of PV inverters (e.g. 50 microinverters at 1.60A each), you'd be better off with a load side connection than a line side connection.

Namely, you use a single 100A service OCPD, so you clearly have a residential service. Now #4s are allowed to be protected at 100A everywhere downstream of that. The 100A service OCPD supplies a #4 feeder, which is spliced to #4 PV conductors and #4 load conductors. The PV conductors go to the combiner panel, and the load conductors go to a main breaker load panel. This complies with the feeder interconnection rules and all is copacetic.

You might say "wait, you've effectively used the 83% rule on the PV side." To which I would say "sure, but it's not a problem. The continuous load is only 80A, and the #4 Cu is good for 85A continuous at 75C." As to being able to protect that #4 Cu at 100A, that's just a weirdness of residential services. The argument does rely on the idea that 310.12 takes precedence over 690.8, though.

Cheers, Wayne

 

[jaggedben]

I'd argue against any application of 310.12 where the inverter output would otherwise require the conductors to be larger. My logic being that whatever load diversity idea applies to single phase conductors feeding a dwelling, it does not apply to PV systems.

That said, I'm not 100% sure, but because I've never really understood the logic behind 310.12, and because I agree with Wayne that there's ambiguity in the code.

 

[wwhitney]

I'd argue against any application of 310.12 where the inverter output would otherwise require the conductors to be larger.

That was my initial position when I first considered this issue a couple years ago. But then I realized that there's no harm in this case (as long as the service OCPD is not 100%-rated), because 83% > 1/125% = 80%. I.e. the conductor's actual ampacity (a continuous rating) will still be greater than the continuous current.

The only reason that conductors need to be upsized for continuous use is to accommodate the limitations of the OCPD, which require the OCPD to be upsized; and then the conductors get upsized to still be adequately protected by that upsized OCPD. So if there's a rule that allows the OCPD to be larger than normal anyway (310.12 in this case), and large enough to comply with the 125% continuous use factor on OCPD sizing, as long as the ampacity of the conductors is greater than the continuous load (or PV supply current), the conductors are fine.

That said, I'm not 100% sure, but because I've never really understood the logic behind 310.12

Two theories. The one I like best is that 310.12 originally only applied to 120/240V 3-wire services, which have only 2 CCCs, and so it was in effect a 1/83% = 1.2 ampacity adjustment factor for having only 2 CCCs. A similar effect can be seen in the ampacity tables for cords in Article 400.

However, the allowance was changed sometime this century to also apply to 120/208V 3-wire services, which have 3 CCCs. That change is not in accordance with this theory, so either the theory is wrong, or the change was a mistake.

The other theory is just "the residential load calcs in 220 are too conservative, so let's apply an additional 83% diversity factor." Except that it would be a mistake to do that in 310 rather than in 220, and there would be no need to change the OCPD/conductor balance with this 83% factor, both the conductor and the OCPD could be reduced.

So take your pick of flawed theories. I think we can conclude this is a not fully rational artifact from past practice.

Cheers, Wayne

 

[jaggedben]

Interestingly, your point about the 125% factor makes sense in many situations, but not where the service conductors land directly on a backfed breaker.

I subscribe to the theory that adding 208V to 310.12 was a mistake.

 

[wwhitney]

Interestingly, your point about the 125% factor makes sense in many situations, but not where the service conductors land directly on a backfed breaker.

Not sure I follow your point, and what distinction you are making, can you elaborate?

Also the distinction between Service Non-Entrance Conductors and Service Entrance Conductors probably bears on this.

Cheers, Wayne

 

[shortcircuit2]

I subscribe to the theory that adding 208V to 310.12 was a mistake.

In Massachusetts we delete the 2nd paragraph of 310.12 which removes the 83% rule to 3-phase. That 2nd paragraph was added in the 2017 NEC.

Some AHJ's here interpret that a tap for 705 sources to SEC sized under the 83% rule, voids the allowance as now those conductors no longer are used only to supply the load of the dwelling.

In 2023...705.11(C)(1) refers to the connection as a tap

In Massachusetts we are on the 2023 NEC and the 2020 NEC 705.11(C) 10', 16.5' and 71' rules are gone.

 

[wwhitney]

Some AHJ's here interpret that a tap for 705 sources to SEC sized under the 83% rule, voids the allowance as now those conductors no longer are used only to supply the load of the dwelling.

That interpretation requires treating the 705 sources as loads. Because the language in 310.12 doesn't use the word only, it just says "supplying the entire load associated with an individual dwelling unit." So I think that's a misinterpretation.

In 2023...705.11(C)(1) refers to the connection as a tap

Well, it uses the phrases "splices and taps" which not the same thing as calling the connection a tap. And my earlier comment was not that "tap" necessarily means "feeder tap" per 240.21(C), just that it often does, which is why I sought clarification.

Cheers, Wayne

 

[jaggedben]

Some AHJ's here interpret that a tap for 705 sources to SEC sized under the 83% rule, voids the allowance as now those conductors no longer are used only to supply the load of the dwelling.

That's a terrible, ignorant, illogical and unfair interpretation. But it's not the first time I've heard something like this and it seems like the language needs to be fixed.

 

[jaggedben]

Not sure I follow your point, and what distinction you are making, can you elaborate?

Also the distinction between Service Non-Entrance Conductors and Service Entrance Conductors probably bears on this.

Cheers, Wayne

What I had in mind was...
If I install a meter/main combo, I bring the service conductors to the meter side and land them on factory supplied lugs on the line side of the meter. There are then a number of factory components between my field installed conductors and the service OCPD. So it stands to reason that my conductors don't need to be upsized for continuous use because of the OCPD.

However if I install a separate meter socket, and field install conductors to a backfed breaker service disconnect in a separate enclosure (or any main breaker panel, actually), then my field installed conductors between meter and panel land directly on the OCPD, then the 125% for continuous use should apply, apparently.

I suppose you will probably say the latter are Service Entrance Conductors. But it's not clear to me that the definitions always will line up with the distinction I'm pointing out. It also leads to the conclusion that perhaps conductors between meter and main panel have to be sized differently than between meter and weatherhead, which I've never seen or heard about.