Being told to upsize service wire due to supply-side connection

[Cprisbrey]

I think the inspector might be considering this from a different angle, and if so, I think he is right. Consider a 50 kW grid-tied PV system line-side connected to a 200A service. The design utilizes five 7.6 kW inverters each landing on a 40A breaker in an AC combiner panel. The inverters have a combined max output current of 160A x 1.25 for continuous current requires 200A OCPD and conductor ampacity. The subject home is a vacation home and only used on the weekends. There will be essentially no load during the summer weekday days, and when the sun is shining, the 1.3 DC:AC ratio will ensure that the system will be exporting 160 amps to the grid for more than 3 hours a day. Seems like the service conductors should be fully rated.

Our system is only putting out 80A with the 125%, on a 200A service. So to me needing to upsize the service, or provide load calcs doesn't make since

 

[ggunn]

Our system is only putting out 80A with the 125%, on a 200A service. So to me needing to upsize the service, or provide load calcs doesn't make since

Are you saying that 125% of the maximum rated current of your inverter(s) is only 80A? It could be that your inspector is thinking that PV current adds to the load current in the service conductors rather than offsetting it; I have run across an inspector or two who thought that. Current would have to be additively flowing in both directions simultaneously for that to be true.

 

[GoldDigger]

I have to agree with the inspector that based entirely on the NEC wording surrounding the reduction in nominal require ampacity for service conductors that "supply the full residential load" you have broken that qualification.
To argue that it at worst reduces the peak load is a logical argument which never works with respect to Code requirements.
It is equally illogical to require the feeder size to increase when, for example, an HVAC load is given a separate connection to the service wires. But we accept that as one of the vagaries of the Code.
The existing service now supplies the full residence load only part of the time.

 

[wwhitney]

I have to agree with the inspector that based entirely on the NEC wording surrounding the reduction in nominal require ampacity for service conductors that "supply the full residential load" you have broken that qualification.

The 2020 NEC 310.12 language is "supplying the entire load associated with" a dwelling unit.

I agree that if that phrase means "carrying all of the energy required by the loads", then once you add a PV line side connection, the service conductors on the grid side of the PV interconnection no longer supply the entire load. [The ones on the house side of the PV interconnection still do.] But that argument would apply equally to a load side PV interconnection, i.e. it would mean 310.12 is inapplicable whether there is interconnected PV.

Since "supplying the entire load" isn't something defined in Article 100, I think we are free to pick an interpretation that gives more reasonable results. Like "capable of carrying energy to all of the loads" associated with a dwelling unit.

Cheers, Wayne

 

[ggunn]

I have to agree with the inspector that based entirely on the NEC wording surrounding the reduction in nominal require ampacity for service conductors that "supply the full residential load" you have broken that qualification.
To argue that it at worst reduces the peak load is a logical argument which never works with respect to Code requirements.

I'm not sure what you are saying. For a given load in the MDP, the service conductors between the tap and the MDP are carrying the same amount of current irrespective of the output of the PV system, but the service conductors between the tap and the transformer supplying the MDP would be carrying the load current minus the PV current, whatever that is.

If the ampacity of the service conductors is sufficient to supply the full load before the addition of the PV, and the service conductors are large enough to carry the maximum PV current if the loads are all off, what has changed? There is no way that the service conductors will be required to carry any more current than they already were.

 

[ggunn]

Why do you see it as a problem? The only problem I see is the requirement to run 250 kcmil Al from the 4/0 Al service conductors to the 200A OCPD. 4/0 Al would be sufficient for those.

The problem I see is that 705.11(A) says "The sum of the power source continuous current output ratings ... shall not exceed the ampacity of the service conductors." For the inverters, the output is 160A X 1.25 = 200A, and with no derating at all the 90 degree ampacity of 4/0 aluminum is only 205A. If the temperature derate drops the ampacity of the 4/0 to less than 200A (200A / 205A = 0.976, which is in the range for 78-86 degrees F), the ampacity is exceeded by the current. The value of the service OCPD is not relevant.

 

[wwhitney]

The problem I see is that 705.11(A) says "The sum of the power source continuous current output ratings ... shall not exceed the ampacity of the service conductors." For the inverters, the output is 160A X 1.25 = 200A.

Nope. The inverters' continuous current output ratings add up to 160A. No 125% factor is specified in 705.11(A). 160A < 180A (75C ampacity of 4/0 Al), 705.11(A) satisfied.

Cheers, Wayne

 

[ggunn]

Nope. The inverters' continuous current output ratings add up to 160A. No 125% factor is specified in 705.11(A). 160A < 180A (75C ampacity of 4/0 Al), 705.11(A) satisfied.

"Continuous output current" does not imply 125%?

 

[wwhitney]

"Continuous output current" does not imply 125%?

Nope. 125% is often (almost always) specified for rules dealing with continuous currents, but it applies only when specified. And the only reason it should be specified in the NEC is for issues related to OCPD performance, and almost all of those rules include exceptions that if the OCPD is 100% rated, the 125% factor does not apply.

Cheers, Wayne

 

[jaggedben]

I have to agree with the inspector that based entirely on the NEC wording surrounding the reduction in nominal require ampacity for service conductors that "supply the full residential load" you have broken that qualification.
To argue that it at worst reduces the peak load is a logical argument which never works with respect to Code requirements.
It is equally illogical to require the feeder size to increase when, for example, an HVAC load is given a separate connection to the service wires. But we accept that as one of the vagaries of the Code.
The existing service now supplies the full residence load only part of the time.

First of all, I think your interpretation of the meaning of 'load' is wrong as I've already explained up thread. (It's VA, not kWh).

But also, your example involving a separate HVAC disconnect is wrong because of 310.12 (C), which exists to exempt the subpanel feeder from being upsized because of the separate disconnect. So don't say we accept such vagaries when, well, we don't. (That is, if we're not misinformed or power mad AHJs).

 

[ggunn]

I still think that it is strange that the PV AC conductors are required to have higher ampacity than the service conductors; if the loads are off the same current runs through both sets of conductors. If that's just an unintended consequence of the NEC rules, then so be it. It isn't alarming to me from a safety perspective but I am willing to bet that there are inspectors who would take issue with it. As I said, I would not do it, but realistically speaking, it's a boundary condition that is not that likely to come up in my world. I have emailed the lead engineer in a major PV design house asking his thoughts on the matter; if he replies to me I will post what he says.

 

[wwhitney]

I still think that it is strange that the PV AC conductors are required to have higher ampacity than the service conductors

Having thought about it some more and reviewed some of Article 230, my current take is that it's in accordance with the allowance in 230.90(A) Exception 3, when there is more than one service OCPD, to protect the service conductors only by load calculation and not directly by the OCPDs.

For example, take a non-residential service with a calculated load of 375A continuous. It could consist of 451A ampacity service (non-entrance) conductors, followed by 469A service entrance conductors, to a 500A OCPD. [230.90(A) says the OCPD has to protect the service conductors while allowing the 240.4(B) round up rule, but 230.42(A) requires the service entrance conductors to have ampacity of 125% of the continuous load.]

If instead there are two service OCPD, with the 375A continuous load split as 325A and 50A, the service (non-entrance) conductors now only need a 375A ampacity. They could supply 2 sets of service entrance conductors, a 406A ampacity set to a 450A OCPD for the 325A continuous load, and a 63A ampacity set to a 70A OCPD for the 50A continuous load.

So now we have another example where the service entrance conductors (comparable to the unprotected PV line side tap conductors) are required to be larger than the service conductors.

Indeed, if the service with 375A continuous load now also has 375A of continuous current inverter outputs, then we could again set it up as two sets of service entrance conductors, one for the loads and one for the PV. In which case the service (non-entrance) conductors need only have a 375A ampacity, while each set of service entrance conductors will need a 469A ampacity.

Cheers, Wayne

 

[ggunn]

It's weird how things work out sometimes; we were just discussing something related to this:

After 4 months of inactivity a client of mine has moments ago come back with some changes he wants in a design I did for him. Originally it was pretty simple - a 600A line side PV connection to a 1200A service - but things have changed. There are presently two service meters at the transformer; one is the 1200A service and the other is a 300A service feeding another building, but the customer is combining them. The branch point for the two existing services is at the transformer, so the POCO is moving their CTs to the transformer side of the branch point to combine the services; so far so good.

There is a gutter next to the transformer where the 300A service CTs used to be which will still have the 300A service conductors passing through it. My client wants to know (and so do I) if he can bring in his 600A PV conductors into that gutter and tap the service conductors if he replaces the service conductors from the tap to the branch point with appropriately sized wiring.

Electrically I don't see a problem; the 300A service conductors already have more than 20,000A of available fault current from the transformer; adding 600A to it doesn't really change anything on the load side of the tap, and on the line side of the tap the conductors will be plenty big.

But will the NEC allow it?

 

[jaggedben]

It's weird how things work out sometimes; we were just discussing something related to this:

After 4 months of inactivity a client of mine has moments ago come back with some changes he wants in a design I did for him. Originally it was pretty simple - a 600A line side PV connection to a 1200A service - but things have changed. There are presently two service meters at the transformer; one is the 1200A service and the other is a 300A service feeding another building, but the customer is combining them. The branch point for the two existing services is at the transformer, so the POCO is moving their CTs to the transformer side of the branch point to combine the services; so far so good.

There is a gutter next to the transformer where the 300A service CTs used to be which will still have the 300A service conductors passing through it. My client wants to know (and so do I) if he can bring in his 600A PV conductors into that gutter and tap the service conductors if he replaces the service conductors from the tap to the branch point with appropriately sized wiring.

Electrically I don't see a problem; the 300A service conductors already have more than 20,000A of available fault current from the transformer; adding 600A to it doesn't really change anything on the load side of the tap, and on the line side of the tap the conductors will be plenty big.

But will the NEC allow it?

I don't see any problem with that.

 

[ggunn]

I don't see any problem with that.

Thanks. Whaddya think, Wayne?

 

[wwhitney]

Thanks. Whaddya think, Wayne?

I certainly think it's fine, it's just like one of my examples. But I'm a self-taught amateur and have no applicable professional experience.

Cheers, Wayne

 

[macmikeman]

All that don't mean squat. The poco is still using a 1/0 alum at best for the supply side conductors. If that connection was on the load side of the service I might agree with the call out. But since it is hooked up line side, it is the power company's decision.

 

[jaggedben]

All that don't mean squat. The poco is still using a 1/0 alum at best for the supply side conductors. If that connection was on the load side of the service I might agree with the call out. But since it is hooked up line side, it is the power company's decision.

We are not talking about poco owned conductors. Look up the meaning of a supply-side connection in 705. I don't think it's what you think.

 

[macmikeman]

I was assuming for supply side connections that meant connected between smarty pants meter and building disconnecting means. So sorry, and carry on .

 

[ggunn]

I was assuming for supply side connections that meant connected between smarty pants meter and building disconnecting means. So sorry, and carry on .

I'm not sure what that "smarty pants" is all about, but yes, we are talking about conductors on the customer's side of the meter ahead of the main service disconnect, which are owned by the customer.