A new one on me

[ggunn]

I have been designing commercial and residential PV systems for 15 years, but this is a new wrinkle for me.

A client has asked me to review a design for a PV system interconnection through a 600A fused disconnect with 500A fuses tapped onto the busbars of a 4000A MDP. The conductors between the tap and the fused disco are two sets of 300kcmil copper. All this stuff is outdoors and the design calls for the disco to be within 10' of the MDP.

705.12(B)(2) specifically talks about connection through a backfed breaker, so I don't think it applies. 705.12(B)(5) may apply, but I don't know yet if the MDP busbars already "supply lugs" (supply is a verb in this context, I believe) or if they plan to add them, in which case does the article apply?

I have never done an interconnection this way, and it looks a bit weird to me. Whaddya think?

 

[pv_n00b]

This is fairly common in larger systems, if they have to be interconnected on the load side with no CB available. Typically a 4,000A SWBD is covered under engineering supervision (705.12(B)(6)). Just check the proposed busbar loading and AIC rating with the PV system added. The conductors between the bus tap and the PV disconnect fall under the tap rules.

 

[ggunn]

This is fairly common in larger systems, if they have to be interconnected on the load side with no CB available. Typically a 4,000A SWBD is covered under engineering supervision (705.12(B)(6)). Just check the proposed busbar loading and AIC rating with the PV system added. The conductors between the bus tap and the PV disconnect fall under the tap rules.

I assume the busbar loading is under 705.12(B)(2) and the tap must be at the opposite end from the MCB.

 

[electrofelon]

How will the busbars be tapped? That could be sticky if the inspector is picky.

 

[ggunn]

How will the busbars be tapped? That could be sticky if the inspector is picky.

That is a question I have indeed brought up with my client.

 

[Fred B]

How will the busbars be tapped? That could be sticky if the inspector is picky.

Wouldn't a panel with feed thru lugs work for that?

 

[electrofelon]

Wouldn't a panel with feed thru lugs work for that?

Maybe you didnt catch it, but this is 4000A, so it will be a switchboard. There is a chance there are some unused lugs just sitting there, but It is unlikely. I have posted this frequently:

https://code-authorities.ul.com/wp-content/uploads/2014/04/ul_TappingBusbarConsiderations.pdf

Essentially any holes in the bus bars need to be labeled with "TAP" to add lugs. You may have a more lax inspector who lets you use any existing holes or add to existing joints without a field evaluation. That happened to me once when I got his approval beforehand and he was ok with me adding lugs to a busbar bolt joint that had a NEMA standard bolt pattern. Another option is to consult the manufacturer, maybe they have some sort of kit available.

 

[jaggedben]

There's no reason you have to conform with 705.12(B)(2) if your asking for approval under 705.12(B)(6). Although if you're the engineer who would stamp it then I see nothing wrong with asking for a connection at the opposite end.

I never really noticed that language about a backfed breaker in 705.12(B)(2) before, but I'd note it only applies to the labeling. The rest of the language is agnostic as to the type of connection. I've always considered that something like a load side side tap to a feeder connected through feed-through lugs at the opposite end could conform with 705.12(B)(2). Perhaps you would change the label if the connection wasn't an 'overcurrent device'.

I don't think that 705.12(B)(5) can apply to your situation because it specifically references panelboards.

 

[electrofelon]

I don't think that 705.12(B)(5) can apply to your situation because it specifically references panelboards.

I am confused now. Which code cycle are we talking about? 705.12(B)(5) applies to everything not just panelboards, at least in the 2020. And I am not seeing the "supplying lugs" wording in 705.12(B)(5). Are we all looking at different code cycles?

 

[jaggedben]

Yes, I'm referencing the 2023. That is where (5) is the one about feed through lugs and (6) is the one about engineering.

They literally swapped the positions of (5) and (6) between the cycles. Obviously for no other reason than to confuse us.

 

[jaggedben]

I guess they had to continue the tradition of remembering something in 705.12 *every code cycle*.

 

[electrofelon]

Ok so anyway, it seems the engineering one is pretty straightforward, biggest challenge might just be how to make the tap itself.

 

[BandGap1.1eV]

Ok so anyway, it seems the engineering one is pretty straightforward, biggest challenge might just be how to make the tap itself.

This is the general consensus I've always encountered. Anything is possible, just costs money. Biggest challenge is getting pricing on that type of system and interconnection when there's this large of an unknown. Then telling the client that we need to shut their whole building down to tear into the gear and figure out the least intrusive way to pull it off. Not many can deal with the amount of risk imposed by that scenario.

Then there's the whole "field listing" topic.

 

[electrofelon]

This is the general consensus I've always encountered. Anything is possible, just costs money. Biggest challenge is getting pricing on that type of system and interconnection when there's this large of an unknown. Then telling the client that we need to shut their whole building down to tear into the gear and figure out the least intrusive way to pull it off. Not many can deal with the amount of risk imposed by that scenario.

Then there's the whole "field listing" topic.

I may very well be simpler to just add a breaker. But it may be worth investigating, it is possible that the bus bar route would be super simple. For example, some inspectors will have a "as long as you don't drill holes" philosophy, you just never know.

 

[ggunn]

Their drawings do not specify the method of making the tap connection, so it is not within the scope of my review, but I have advised the client to take it up with the AHJ to make sure they approve of their methodology.

 

[pv_n00b]

I assume the busbar loading is under 705.12(B)(2) and the tap must be at the opposite end from the MCB.

We do it mostly under engineering supervision, 705.12(B)(6), so not as worried about the other provisions as long as the calculated bus loading is good and AIC ratings are good.
In a large SWBD, like the 4,000A SWBD you are working with, typically finding a good spot to land lugs in existing holes, even a large number of parallel conductor lugs, is not that hard. Now try doing the same thing in a supply side interconnection where there is 10" of bus between the main CB line side and the metering compartment. That's difficult.