Dedicated PV Load Center

[fossil112]

All,
For the life of me, I cannot remember the NEC section that deals with dedicated PV load centers. I have a panelboard with only PV inverters for an off grid system with battery backup. Couldn't remember the rules on sizing the panelboard bus ratings and it's OCPD's. Any help? Referencing the 2011 NEC.

 

[ggunn]

All,
For the life of me, I cannot remember the NEC section that deals with dedicated PV load centers. I have a panelboard with only PV inverters for an off grid system with battery backup. Couldn't remember the rules on sizing the panelboard bus ratings and it's OCPD's. Any help? Referencing the 2011 NEC.

There is no article in the code which specifically addresses this, and it has been a subject of some debate in here as to if and how the 120% rule applies. I'll hang up and listen.

 

[Smart $]

There is no article in the code which specifically addresses this, and it has been a subject of some debate in here as to if and how the 120% rule applies. I'll hang up and listen.

AFAICT the 120% rule only applies to "utility-interactive inverters", which is the title of 705.12(D).

 

[ggunn]

AFAICT the 120% rule only applies to "utility-interactive inverters", which is the title of 705.12(D).

Why would that be? I mean, I can understand the distinction in that the main breaker cannot feed the AC combiner in an off grid system, but it seems to me that you still wouldn't want to feed a 100A panel/busbar with more than 120A of backfed breakers.

 

[Smart $]

All,
For the life of me, I cannot remember the NEC section that deals with dedicated PV load centers. I have a panelboard with only PV inverters for an off grid system with battery backup. Couldn't remember the rules on sizing the panelboard bus ratings and it's OCPD's. Any help? Referencing the 2011 NEC.

Seems 690.10 applies, but says little in this regard. The majority of applicable requirements are in Chapter 2.

 

[Smart $]

Why would that be? I mean, I can understand the distinction in that the main breaker cannot feed the AC combiner in an off grid system, but it seems to me that you still wouldn't want to feed a 100A panel/busbar with more than 120A of backfed breakers.

True. On the surface, its actually worse than the 120% rule. The busbar may have to be rated for 100% of the backfed breakers sum. However, we can apply tap rules to the busbar... the load and load breakers will determine the outcome. FWIW, 690.10(A) says the AC output can be less than the calculated load.

 

[Smart $]

... Any help? ...

Perhaps you can provide some more details (inverter number and output ratings, load ratings) and we can [try to] help you

 

[fossil112]

I know we're going to have about 15 PV inverters on a single load center, probably somewhere around 300kWAC...details make no difference. I know someone out there remembers the little blurb...maybe it's an informational note that I've missed.

 

[SolarPro]

Maybe you are looking for the Exception to 705.12(D)(2), added in NEC 2011.

 

[Smart $]

I know we're going to have about 15 PV inverters on a single load center, probably somewhere around 300kWAC...details make no difference. I know someone out there remembers the little blurb...maybe it's an informational note that I've missed.

My non-assistance will make no difference, too.

 

[tallgirl]

Seems 690.10 applies, but says little in this regard. The majority of applicable requirements are in Chapter 2.

690.10 says more than plenty enough. The same rules as if the standalone system were connected to a service apply. The differences from "everything else in the Code" are listed as the exceptions A through E.

 

[Smart $]

690.10 says more than plenty enough. The same rules as if the standalone system were connected to a service apply. The differences from "everything else in the Code" are listed as the exceptions A through E.

Well not exactly... and depending on one's perspective, nowhere close. An SAPVS is only required to supply enough power for the largest utlization load, whereas a service must be sized to supply all loads. And how often do you have 15 service disconnects supplying one panel...???

 

[tallgirl]

Well not exactly... and depending on one's perspective, nowhere close. An SAPVS is only required to supply enough power for the largest utlization load, whereas a service must be sized to supply all loads. And how often do you have 15 service disconnects supplying one panel...???

1). I quoted that from the NEC, which is sitting on my desk at the office.
2). A service is NOT required to be sized to supply all loads.
3). You need a new engineer if someone is designing systems with 15 disconnects supplying on panel.

 

[Smart $]

1). I quoted that from the NEC, which is sitting on my desk at the office.
2). A service is NOT required to be sized to supply all loads.
3). You need a new engineer if someone is designing systems with 15 disconnects supplying on panel.

1. The location of your copy of the NEC is irrelevant, regardless of where you are. And at best you paraphrased. Here's an actual quote:

690.10 Stand-Alone Systems. The premises wiring system
shall be adequate to meet the requirements of this Code
for a similar installation connected to a service. The wiring
on the supply side of the building or structure disconnecting
means shall comply with this Code except as modified
by 690.10(A) through (E).

2. A service IS required to be sized to supply ALL connected loads, just not necessarily all at the same instance. An SAPVS does not have to match that requirement.

3. Your opinion is noted as such. However, my opnion is 15 PV inverter disconnects supplying one panel is not beyond the scope of the NEC, nor does it stand out as a bad design.

 

[tallgirl]

1. The location of your copy of the NEC is irrelevant, regardless of where you are. And at best you paraphrased. Here's an actual quote:

​

2. A service IS required to be sized to supply ALL connected loads, just not necessarily all at the same instance. An SAPVS does not have to match that requirement.

3. Your opinion is noted as such. However, my opnion is 15 PV inverter disconnects supplying one panel is not beyond the scope of the NEC, nor does it stand out as a bad design.

I'd suggest that a service that doesn't have to be sized to supply "ALL connected loads, just not necessarily all at the same instance" is not, in fact, sized to supply "ALL" connected loads.

And I'll stick with my opinion that having 15 inverter disconnects supplying a single panel is a miserable design and the person should be introduced to aggregation as a concept. If what you really mean is that you have a main breaker aggregation panel which is then supplying a panel, I'll take back what I said. But if what you have is a panel, with its connected loads, and 15 inverters, which have their disconnects attached directly to that same panel ("15 PV inverter disconnects supplying one panel") I'm going to stick with my professional opinion (since I'm a PV professional ...) that you need a new engineer.

 

[Smart $]

I'd suggest that a service that doesn't have to be sized to supply "ALL connected loads, just not necessarily all at the same instance" is not, in fact, sized to supply "ALL" connected loads.

Contrasting an SAPVS as only required to be sized to supply the largest single utilization equipment, the distinction is mere semantics.

... If what you really mean is that you have a main breaker aggregation panel which is then supplying a panel, I'll take back what I said. ...

I believe you'd have to meet the six-disconnect rule... but depending on breaker poles, handle ties are not beyond reproach. Six three-handle ties could disconnect 18 1-pole breakers. I'm not saying I would do this, but appears to be permitted under Code.

 

[jaggedben]

I believe you'd have to meet the six-disconnect rule... but depending on breaker poles, handle ties are not beyond reproach. Six three-handle ties could disconnect 18 1-pole breakers. I'm not saying I would do this, but appears to be permitted under Code.

It's not clear to me that the six-disconnect rule (690.14(B)(4)) applies to disconnects for inverters. That's because it's not clear to me that the code defines multiple inverters on the same premises as one system. I think 690.14(B)(4) applies primarily to DC disconnects. And if all inverters are aggregated to a single main breaker, I'd say that's one handle.

 

[Smart $]

It's not clear to me that the six-disconnect rule (690.14(B)(4)) applies to disconnects for inverters. That's because it's not clear to me that the code defines multiple inverters on the same premises as one system. I think 690.14(B)(4) applies primarily to DC disconnects. And if all inverters are aggregated to a single main breaker, I'd say that's one handle.

I agree... its not clear... especially considering the wording of 690.15

However, my reference to the six-disconnect rule is based on...

690.10 Stand-Alone Systems. The premises wiring system
shall be adequate to meet the requirements of this Code
for a similar installation connected to a service. The wiring
on the supply side of the building or structure disconnecting
means shall comply with this Code except as modified
by 690.10(A) through (E).

...and the six-disonnect rule applies to a premises wiring system supplied by a service.

 

[tallgirl]

Contrasting an SAPVS as only required to be sized to supply the largest single utilization equipment, the distinction is mere semantics.


I believe you'd have to meet the six-disconnect rule... but depending on breaker poles, handle ties are not beyond reproach. Six three-handle ties could disconnect 18 1-pole breakers. I'm not saying I would do this, but appears to be permitted under Code.

I still think a better design, and what I've seen used on all the PV projects I've worked on with more than 1 or 2 inverters per pole, is to aggregate the inverters, each with their own disconnect, and have a disconnect for the aggregated output.

 

[Smart $]

I still think a better design, and what I've seen used on all the PV projects I've worked on with more than 1 or 2 inverters per pole, is to aggregate the inverters, each with their own disconnect, and have a disconnect for the aggregated output.

I agree... it is a better design. But it is not required by Code. The downside, though many think otherwise, is the "120% rule" then applies, even though there is no branch-circuit load connected to the aggregation panel (except when all available spaces are used disconnecting inverters).