Here's one for you...

[ggunn]

We have a project on the roof of a new condo complex that is 100+ small PV systems each feeding a single condo unit. Each unit has its own meter and 100A service disco feeding a 100A MLO MDP in one of three ground floor electrical rooms, and each MDP has a 100A load breaker that feeds a panel in the unit on an upper floor. Our systems feed the MDPs in the electrical rooms through 20A breakers. We feed the PV at the opposite end from the utility feed, the bus is qualified under 705.12(D)(2)(3)(b), i.e., the 120% rule, and everything's hunky, right?

But no; the EC installed MDPs with a weird bus that feeds L1 from one end and L2 from the other. I've never seen this but our master says they are not that uncommon. I don't see how we can qualify under any of 705.12(D)(2)(3), not even (d) since the bus is neither center fed nor multiple ampacities.

We have investigated downsizing the breakers feeding the units to 80A in order to qualify the bus under (c) but the unit load analysis the EC ran shows that we can't do that. The meter and service disco are in the same enclosure, so we can't get to the line side of the disco. The MDPs are MLO so we can't install load side taps between them and the service discos. I don't think we can interconnect via two single pole breakers each on the opposite end of the bus from the utility feed.

Whaddya think?

 

[SolarAlternativesDesign]

How about a Renewable Meter Adapter like this:
http://connectder.com/products
Not every utility would accept these, but they are UL listed.

 

[ggunn]

How about a Renewable Meter Adapter like this:
http://connectder.com/products
Not every utility would accept these, but they are UL listed.

I have searched the web but I cannot find even a ballpark number for what these gizmos cost. Have you any idea?

 

[SolarAlternativesDesign]

I got an email from connectDER in January 2017 saying that the 20A - 60A size was $400 and the 70A - 80A size was $450. They said they sell direct. PG&E does not approve these. SDG&E has their own version which they install themselves for $1,326 last January.

 

[ggunn]

I have searched the web but I cannot find even a ballpark number for what these gizmos cost. Have you any idea?

Never mind; I got a quote from ConnectDER: $449, discounted to ~$350 in bulk. For ~100 very small PV systems it is cost prohibitive.

EDIT: I just saw your response; it looks like they quoted me the price for the larger unit, but even at $50 less it's still way too expensive.

 

[jaggedben]

You could tap the feeders if you could somehow install main breakers in the MDPs.

I gather that virtual net metering is not an option.

 

[ggunn]

You could tap the feeders if you could somehow install main breakers in the MDPs.

I gather that virtual net metering is not an option.

We thought of that, too, but I don't know of a main breaker that can feed buses with lugs on the opposite ends from each other. Virtual metering isn't an option, either. There was a communication breakdown; the original electrical drawings for the building showed 125A buses, but to save space the EC changed them to 100A. The asked us and we OK'ed it; our breakers are all 20A, so the 120% rule would normally apply, but they didn't tell us about the weird bus configuration. If they were 125A we could have qualified under (a) or (c) even with the weird buses.

We're looking for options; I don't think we can justify interconnecting with a pair of 1P breakers, but can you give me a code reference that prohibits it?

 

[SolarAlternativesDesign]

You could tap the feeders if you could somehow install main breakers in the MDPs.

I gather that virtual net metering is not an option.

Would installing main breakers in the MDP and a tap void the panel listing? Could you explain how this would work (in a conventional panel, not the one ggunn is working with)?

 

[ggunn]

Would installing main breakers in the MDP and a tap void the panel listing? Could you explain how this would work (in a conventional panel, not the one ggunn is working with)?

Some of the MLO panels we encounter are convertible, i.e., you can install main breaker kits in them and their deadfronts have knockouts to accommodate the breakers. Others are not, but sometimes you can add a tied down backfed load breaker to accomplish the same thing. Unfortunately, the panels I am dealing with now do not allow for either modification. Adding a "tap" (it's not really a tap under the code, but anyway...) inside a panel enclosure does not void its listing in the view of all but one of the AHJs I deal with.

 

[jaggedben]

We thought of that, too, but I don't know of a main breaker that can feed buses with lugs on the opposite ends from each other. Virtual metering isn't an option, either. There was a communication breakdown; the original electrical drawings for the building showed 125A buses, but to save space the EC changed them to 100A. The asked us and we OK'ed it; our breakers are all 20A, so the 120% rule would normally apply, but they didn't tell us about the weird bus configuration. If they were 125A we could have qualified under (a) or (c) even with the weird buses.

I'd take a guess that your MDPs are Square D HOM6-12L100SCP or similar, in which case there is a provision for a main breaker with hold down, to my knowledge. That would turn them into an end-fed panel as well. However it would also use up 4 of the 12 spots for branch circuits so that could be a problem of space. It would only work if the condo branch circuits number 8 poles or fewer. If they are 8 you could tap, if they are 6 you have rescued your original plan for about $50 per connection.

I gather that 'change out all the MDPs, or add new subs for space' is not solution that excites you.

We're looking for options; I don't think we can justify interconnecting with a pair of 1P breakers, but can you give me a code reference that prohibits it?

Not off the top of my head but it would be in the handle tie provisions for MWBCs.

Would installing main breakers in the MDP and a tap void the panel listing? Could you explain how this would work (in a conventional panel, not the one ggunn is working with)?

What's to explain? If the panelboard has a provision for a main then you install the main and tap the feeder ahead of it with polaris or insulation piercing connectors. The busbar arrangment is irrelevant because it's downstream.

 

[ggunn]

I'd take a guess that your MDPs are Square D HOM6-12L100SCP or similar, in which case there is a provision for a main breaker with hold down, to my knowledge. That would turn them into an end-fed panel as well. However it would also use up 4 of the 12 spots for branch circuits so that could be a problem of space. It would only work if the condo branch circuits number 8 poles or fewer. If they are 8 you could tap, if they are 6 you have rescued your original plan for about $50 per connection.

I gather that 'change out all the MDPs, or add new subs for space' is not solution that excites you.

Unfortunately, there is no main breaker kit that we have been able to find that would convert the panels into main breaker panels, plus the enclosures the EC is installing are so small that there is not enough room in them to do much if any modification. The MDPs are not in our scope of work; if they were we would not have done this to ourselves. The EC's original plans called for 125A panels; if they had stuck with that we wouldn't have had a problem. They asked us if switching to 100A panels would work for us and we told them it was OK, but they didn't tell us that the panels would have this strange (to me, anyway) bus configuration.

What we have heah <dramatic pause> is a failure to communicate.

 

[wwhitney]

A few comments/questions (based on 2017 NEC), not sure if any of them will actually help:

1) This 100A MLO MDP you refer to, the design was to have only two breakers in it? A 20A breaker for the solar interconnection and a 100A breaker for the feeder to the unit? That's my assumption for the rest of the comments.

2) I assume interconnecting in each unit is not an option.

3) I think there is a reasonable argument that 705.12(B)(2)(b) and (d) should together cover all panels. I.e. since "center fed" is not defined, it could be taken to mean "fed in any way other than from one end." Then 705.12(B)(2)(d) would apply to your situation.

4) I don't understand why 705.12(B)(2)(c) counts both load and supply breakers; it seems to me that if the sum of all load breakers is less than the busbar rating, that ought to be sufficient protection for the busbar. Of course, that doesn't help you.

5) Maybe argue for treating the interconnection under 705.12(B)(2)(1), viewing the panel as just a point of interconnection to the 100A feeder from the service entrance to the distribution panel in the unit.

Cheers, Wayne

 

[wwhitney]

So, in case it got lost in my longer post, I still think that for the 2017 NEC, at least, the intention is likely that every panel (with a properly located interconnection breaker) should fall under 705.12(D)(2)(3)(b) or (d). In which case since your panel doesn't fall under (b), section (d) should cover it. One could check the ROP et al to see if that was the intention.

Also, what about replacing the 100A breaker in the MDP that feeds the unit with 100A subfeed lugs? I think that would work if there is a 100A main breaker in the unit's distribution panel. There's already a 100A breaker in the service disconnect, as I understand it, that would be protecting the feeder as far as the primary supply. And the feeder would be protected as far as the utility interactive inverter under 705.12(D)(2)(1)(b). The MDP would qualify under 705.12(D)(2)(3)(c), unless there are several other breakers in it that you haven't mentioned.

Cheers, Wayne

 

[pv_n00b]

We're looking for options; I don't think we can justify interconnecting with a pair of 1P breakers, but can you give me a code reference that prohibits it?

Check out NEC 240.15, it seems to be saying that using (2) 1P breakers on a 240V circuit would not be allowed. A fault on either conductor has to open both legs.

 

[pv_n00b]

We have a project on the roof of a new condo complex that is 100+ small PV systems each feeding a single condo unit. Each unit has its own meter and 100A service disco feeding a 100A MLO MDP in one of three ground floor electrical rooms, and each MDP has a 100A load breaker that feeds a panel in the unit on an upper floor. Our systems feed the MDPs in the electrical rooms through 20A breakers. We feed the PV at the opposite end from the utility feed, the bus is qualified under 705.12(D)(2)(3)(b), i.e., the 120% rule, and everything's hunky, right?

But no; the EC installed MDPs with a weird bus that feeds L1 from one end and L2 from the other. I've never seen this but our master says they are not that uncommon. I don't see how we can qualify under any of 705.12(D)(2)(3), not even (d) since the bus is neither center fed nor multiple ampacities.

We have investigated downsizing the breakers feeding the units to 80A in order to qualify the bus under (c) but the unit load analysis the EC ran shows that we can't do that. The meter and service disco are in the same enclosure, so we can't get to the line side of the disco. The MDPs are MLO so we can't install load side taps between them and the service discos. I don't think we can interconnect via two single pole breakers each on the opposite end of the bus from the utility feed.

Whaddya think?

That’s a tough nut. The NEC does not cover this situation so you either have to not touch the MDPs or work out an acceptable alternative means and methods with the AHJ. Conceptually it’s the same as a center fed bus. If you put the PV breaker on one end of a center fed bus half the bus is protected from overload and the other half is not, same situation you have with this MDP. If you draw both out and compare you might have a reasonable chance of selling it to the AHJ.

 

[petersonra]

I feel your pain. I think you are screwed. It seems to me that to get a permit for something like this some kind of plan sketch or drawing with or without an engineer seal on it had to have been submitted to get a permit. If so how did it come to be that somebody accepted this kind of change? If your answer is that the EC is acting as design and build that it seems to me that the EC has to accept responsibility for making this right and not trying to con the building department into accepting something that does not meet code.

 

[ggunn]

I feel your pain. I think you are screwed. It seems to me that to get a permit for something like this some kind of plan sketch or drawing with or without an engineer seal on it had to have been submitted to get a permit. If so how did it come to be that somebody accepted this kind of change? If your answer is that the EC is acting as design and build that it seems to me that the EC has to accept responsibility for making this right and not trying to con the building department into accepting something that does not meet code.

It's resolved. I met with the AHJ and invoked 705.12(D)(2)(3)(d); although these panels are not technically center fed, I maintained that they are similar enough in topology to be covered. Since the only pathway out of the panel is through the 100A load breaker, and with the caveat that no other load breakers shall be installed in the panel (the EC was on site and confirmed this and we will label the panel accordingly), there is no way to overload the bus. The AHJ concurred.

 

[SolarAlternativesDesign]

If the panelboard has a provision for a main then you install the main and tap the feeder ahead of it with polaris or insulation piercing connectors. The busbar arrangment is irrelevant because it's downstream.

We haven't had experience with that - just trying to learn. Sounds like tapping the feeder ahead of the main with insulation piercing connectors would be a simple method if there is enough room. If installing Polaris connectors, would that modify the feeder connection to the main so much that the panel listing or warranty would be jeopardized? When I talked with Eaton tech support about replacing a feeder above the main breaker (to make room for installing a CT) he said Eaton would approve disconnecting and reconnecting the feeder but not replacing it with a longer one - hadn't been tested. Maybe a panelboard that has a provision for adding a main is a different case.
Thanks.

 

[SolarAlternativesDesign]

Great solution, ggunn.

 

[jaggedben]

We haven't had experience with that - just trying to learn. Sounds like tapping the feeder ahead of the main with insulation piercing connectors would be a simple method if there is enough room. If installing Polaris connectors, would that modify the feeder connection to the main so much that the panel listing or warranty would be jeopardized? When I talked with Eaton tech support about replacing a feeder above the main breaker (to make room for installing a CT) he said Eaton would approve disconnecting and reconnecting the feeder but not replacing it with a longer one - hadn't been tested. Maybe a panelboard that has a provision for adding a main is a different case.
Thanks.

Why would a panelboard manufacturer have any real say in the length of a field installed feeder wire? The NEC allows splices in panelboard enclosures up to certain fill limits. Adding splices to field installed conductors has just about nothing to do with the panelboard listing. The exception might be when it comes to sectional switch gear, and what you are allowed to put in certain sections, I've seen that come up. But I've never had an AHJ question this sort of thing in a small amperage panelboard enclosure. Not on the basis of any listing that is.

Glad ggunn found a simpler solution though.