interconnection to integrated meter and service panel

[jsherwin]

I have a scenario where an installer wants to interconnect a pv inverter to an integrated meter / service panel with 4 spaces available for load breakers and feed through lugs at the bottom. The ocpd for the integrated meter/service panel is 150 amps (which is the busbar rating as well) and it feeds a 150 amp load center (with 150 amp ocpd) via the feed through lugs . The inverter Imax is 32 amps and ocpd is 40 amps. The 40 amp inverter breaker will go in spaces between the 150 amp main ocpd and the 150 amp load center ocpd. A label will be applied to the integrated panel saying that no additional breakers can be installed.

Question: This connection cannot be made under 705.12(D)(2)(3)(a)-(c) however, the busbar on the integrated panel will never see more than 150 amps per the breakers at the meter and load center. Essentially, the busbar is a 150 amp rated feeder to the load center in the garage. By putting the 40 amp ocpd in the empty slots, you are essentially making a "tap" on that feeder. Therefore, interconnecting the inverter to the empty slots in that integrated panel would be nothing more that a tap connection. Is this a code compliant interconnection despite the 705.12(D)(2)(3)(a)-(c) rules?

 

[jaggedben]

That's kind of a new one.

The feed through lugs are a bit of a red herring. The problem is that you can't meet any of the rules for a panelboard. One way to change that could be to downsize the main breaker to 125A.

Or, with the placard you mentioned, you could ask for an alternative methods or materials exception under 90. Since as you pointed out the load will never exceed the busbar, it's pretty certain it's safe.

 

[jsherwin]

all in one - integrated meter meter can and service pane revisited

all in one - integrated meter meter can and service pane revisited

[FONT=&quot]Another twist to the question initially posted. These "all in one" cans [/FONT][FONT=&quot]do not allow for a supply side tap, so what is happening is [/FONT]
[FONT=&quot]that installers are "tapping" into the feeder coming out of the lugs at the bottom [/FONT]
[FONT=&quot]of the panel as it goes to a subpanel elsewhere. This occurs when the [/FONT]
[FONT=&quot]systems are to large and they cannot apply 705.12(D)(2)(3)(a-c) and can't make a supply side connection, so....they [/FONT]
[FONT=&quot]just go past the bus bar and tap into the feeder. They want to call this a "tap" connection, when they really only have moved down a few inches off the buss and onto the feeder wire. Seems like this is very [/FONT]
[FONT=&quot]"gray". What do you think? and how do you guys deal with this? I see [/FONT]
[FONT=&quot]these type of panels all over so I am wondering how an interconnection [/FONT]
[FONT=&quot]can be made without replacing the service, downsizing the main OCPD, or [/FONT]
[FONT=Lucida Grande, Verdana, Arial, Helvetica, sans-serif]reducing system size?[/FONT]

 

[jaggedben]

Another twist to the question initially posted. These "all in one" cans do not allow for a supply side tap, so what is happening is
that installers are "tapping" into the feeder coming out of the lugs at the bottom
of the panel as it goes to a subpanel elsewhere. This occurs when the
systems are to large and they cannot apply 705.12(D)(2)(3)(a-c) and can't make a supply side connection, so....they
just go past the bus bar and tap into the feeder. They want to call this a "tap" connection, when they really only have moved down a few inches off the buss and onto the feeder wire. Seems like this is very
"gray". What do you think? and how do you guys deal with this? I see
these type of panels all over so I am wondering how an interconnection
can be made without replacing the service, downsizing the main OCPD, or
reducing system size?

Take a look at 705.12(D)(2)(2). The code provides rules for taps. Basically it's fine if they can also add a main breaker to the subpanel. However, this doesn't get around rules in the main panel. Since the inverter connection to the main panel is at the feed through lugs at the opposite end, the main panel must comply with 705.12(D)(2)(3)(b), the traditional '120% rule.' If this is the same service as the one you described above, a main breaker downsize will still be required. Downsizing the main breaker is ordinarily not a deal breaker.

 

[jaggedben]

Looking at your description again, installing the inverter breaker to the busbar seems to comply with the letter of 705.12(D)(2)(3)(c), if not with the spirit. (There will probably be a change to the 2020 NEC that addresses feed through lugs and adds additional requirements.) If you also require the downstream subpanel to have a main breaker per 705.12(D)(2)(2), the installation would be safe.

 

[jsherwin]

Thank you for your reply. However, I am still having trouble getting my head around this. Look at image below:
200A MOCPD
200A BUS
feeder ampacity 200A
200A subpanel at end of feeder w/200A ocpd

50A pv/inverter output

Looking at the breaker section of the panel below:

705.12(D)(2)(3)(a) - no (50+200>200)
705.12(D)(2)(3)(b) - no (50+200>240)
705.12(D)(2)(3)(a) - no (50+200>200)
I cannot make a loadside connection on the bus with a breaker.
however,
If I go down to the feeder wires going out of the breaker section, I can tap into that feeder per 705.12(D)(2)(2) and get my 50 amps of pv in?

 

[jaggedben]

Read 705.12(D)(2)(3)(c) again. It says the sum of overcurrent devices excluding the primary overcurrent device. So that sum is just the PV breaker, thus 50A. Again, I admit this reading follows the letter but not the spirit l, since the feed through lugs represent a potential 200A load.

 

[ggunn]

Read 705.12(D)(2)(3)(c) again. It says the sum of overcurrent devices excluding the primary overcurrent device. So that sum is just the PV breaker, thus 50A. Again, I admit this reading follows the letter but not the spirit l, since the feed through lugs represent a potential 200A load.

The AHJ may interpret this section of code such that whatever load breaker(s) are on the feed through conductors are also connected to the bus. I've got a project proposed to an AHJ plan review that hinges on this point; I'll try to remember to post how it comes out, though if it passes the review I may not get notified. If it fails I'll definitely know.

 

[jsherwin]

Read 705.12(D)(2)(3)(c) again. It says the sum of overcurrent devices excluding the primary overcurrent device. So that sum is just the PV breaker, thus 50A. Again, I admit this reading follows the letter but not the spirit l, since the feed through lugs represent a potential 200A load.

Jaggedben,
The 200 in the 705.12(D)(2)(3)(c) calc above represents the 200a (load) subpanel. Despite that breaker not being located directly on that main panel, it would still be considered a load and therefore used in (c). I apologize for the confusion (ie 200a mocp = 200a sub ocpd).

I have reached out to a well respected installer (NABCEP Pro, PE, CEC) who has used this connection method under 705.12(D)(2)(1 and 2).
At the end of the day, the argument being made is that with the 200a mocpd protecting the 200a bus and feeder along with the 200a sup panel mocpd located at the end of a 200a sized feeder, you cannot overload anything.
If the subpanel could pull all of 200 amps and the inverter was putting out a full 60 then, only 140 would come from utility. In the case where the sub used nothing, the inverter would just put 60 back on grid.
The caveat being that no loads would be permitted on the main panel section.

 

[ggunn]

Jaggedben,
The 200 in the 705.12(D)(2)(3)(c) calc above represents the 200a (load) subpanel. Despite that breaker not being located directly on that main panel, it would still be considered a load and therefore used in (c). I apologize for the confusion (ie 200a mocp = 200a sub ocpd).

I have reached out to a well respected installer (NABCEP Pro, PE, CEC) who has used this connection method under 705.12(D)(2)(1 and 2).
At the end of the day, the argument being made is that with the 200a mocpd protecting the 200a bus and feeder along with the 200a sup panel mocpd located at the end of a 200a sized feeder, you cannot overload anything.
If the subpanel could pull all of 200 amps and the inverter was putting out a full 60 then, only 140 would come from utility. In the case where the sub used nothing, the inverter would just put 60 back on grid.
The caveat being that no loads would be permitted on the main panel section.

I tried to explain all that to a particular AHJ chief inspector, but he insisted that the feeder be sized to the main breaker in the sub plus 125% of the nameplate current rating of the inverter(s). Pointing to 705 was to no avail. Appealing to his (lack of) understanding of basic electricity only got him angry. He even managed to get his requirement written into the local interconnection guide which was ratified by the city council and therefore carries the force of law, so three engineers who work for the AHJ, although they all agreed with my analysis, refused to put it in writing.

We don't interconnect PV in subpanels in his jurisdiction.

 

[jaggedben]

Jaggedben,
The 200 in the 705.12(D)(2)(3)(c) calc above represents the 200a (load) subpanel. Despite that breaker not being located directly on that main panel, it would still be considered a load and therefore used in (c). I apologize for the confusion (ie 200a mocp = 200a sub ocpd).

I have reached out to a well respected installer (NABCEP Pro, PE, CEC) who has used this connection method under 705.12(D)(2)(1 and 2).
At the end of the day, the argument being made is that with the 200a mocpd protecting the 200a bus and feeder along with the 200a sup panel mocpd located at the end of a 200a sized feeder, you cannot overload anything.
If the subpanel could pull all of 200 amps and the inverter was putting out a full 60 then, only 140 would come from utility. In the case where the sub used nothing, the inverter would just put 60 back on grid.
The caveat being that no loads would be permitted on the main panel section.

The actual language of 705.12(D)(2)(3)(c) refers to "...all overcurrent devices on panelboards...". It's debatable if the remote subpanel breaker is 'on' the panelboard. Also there are some nuances to the rules behind the code that have to do with busbar heating and not just Kirchoff's law. Do lugs have a different effect on busbar heating than plug on breakers? I don't know, but perhaps.

Anyway...

I think we're seeing eye to eye on what the issues are here, and I think that if you have a warning label prohibiting adding loads to the panelboard then the installation is as safe as any other option under 705.12(D)(2)(3) which requires warning labels.

 

[PWDickerson]

I have run into this situation several times, and there is a gap in the code here that I hope is addressed in the next revision. The proposed installation certainly meets the letter of 705.12(D)(2)(3)(c). I don't think anyone here can argue with that. The service equipment has no load breakers, and so one should be able to backfeed that panel with up to 125A of continuous inverter current on a 150A breaker if the breaker stabs are rated for it. This particular arrangement would also meet the spirit of the code as there is no way to overload any of the busbars or feeders.

However, when there are load breakers in the service panel, one can arrange the equipment in a way that could be unsafe and still meet the letter of the code. The way to do it safely would be to make sure there is a main breaker in the sub-panel, and to locate the backfeed breaker opposite the main breaker in the service panel. Of course, 705.12(D)(2)(3)(c) doesn't talk about backfeed breaker location, but it is important to consider when you have feed through-lugs. If you had a load breaker between the backfeed breaker and the feed-through lugs, the section of busbar between the backfeed and load breakers could see current up to the sum of the main and backfeed breaker ratings, which would overload the busbar. With the backfeed breaker opposite the main, you can't overload anything.

If the OP's installer chose to backfeed at a tap on the feeders to the sub-panel, he/she could safely install up to 125A of continuous inverter current on a 150A OCPD. The service equipment busbar would be qualified under 705.12(D)(2)(3)(c) with the appropriate label. This would meet the letter and spirit of the code.

 

[jaggedben]

The first draft of the 2020 NEC contained a new section on feed-through lugs that read as follows:

(f) Connections shall be permitted on busbars that supply feed-through lugs and conductors
connected to the lugs opposite the main source of supply. The ampacity of the busbar and
connected feeders shall not be less than the sum of the primary source overcurrent device and 125
percent of the power-source output circuit current.

This is a more restrictive rule. There were two public inputs to the second draft that suggested revisions to this language, so it may very well change.

 

[jaggedben]

... Of course, 705.12(D)(2)(3)(c) doesn't talk about backfeed breaker location, but it is important to consider when you have feed through-lugs. If you had a load breaker between the backfeed breaker and the feed-through lugs, the section of busbar between the backfeed and load breakers could see current up to the sum of the main and backfeed breaker ratings, which would overload the busbar. With the backfeed breaker opposite the main, you can't overload anything....

I don't really follow. If the feed through lugs are at the opposite end, as they often are, then when putting the source breaker at the closest breaker location to the opposite end, you still have a section of busbar between the breaker and the lugs that can see an overload.

 

[PWDickerson]

As long as you have a main breaker in the sub-panel, you won't get excess current coming off the end of the main panel bus and feeder conductors.