# Determining maximun inverter size on Feeder connection

#### henryo

##### Member
Hi there,

Need your your help in determining the maximum size of inverter (in AC amp rating will do) if the connection is done on a feeder.

I know there there is a 120% rule and a 100% rule if connections are doen on a buss bar or on the electrical panel butI am not sure how to determine is the connection is on the feeder.

I have attached an illustration to show my point.

#### Attachments

• 392.9 KB Views: 13

#### Carultch

##### Senior Member
Hi there,

Need your your help in determining the maximum size of inverter (in AC amp rating will do) if the connection is done on a feeder.

I know there there is a 120% rule and a 100% rule if connections are doen on a buss bar or on the electrical panel butI am not sure how to determine is the connection is on the feeder.

I have attached an illustration to show my point.

For this particular setup, you also have the 120% rule that applies on the 1200A panelboard. You apply the 120% rule on all panels from the point of interconnection, all the way up to the service point, which is why you usually want to interconnect on the main panel instead of subpanels. In NEC2014 and later, this would limit you to 80% of 240A, i.e. 192A. That would imply up to a 250A OCPD for the interconnection tap circuit, but the total operating current of the source would be limited to 80% of 240A instead of 80% of 250A.

If upstream equipment doesn't have its own rules that otherwise limit you, the rule for interconnecting on feeders in general is as follows. There are 3 cases, depending on whether an OCPD is at both ends of the feeder to protect from potential overload, and depending on whether 240.4(B) was applied when sizing the feeder. Typical of all 3 cases, the short section of conductor between the feeder and the first PV-specific OCPD would be constructed to comply with one of the 240.21(B) tap rules. Maintain symmetry when parallel conductors are involved.

Case 1: There is an OCPD at both ends of the feeder, and feeder conductors meet or exceed the OCPD protecting them (i.e. it sizing is not taking credit for 240.4(B)):
You can connect anything you can fit behind an OCPD, up to the full rating of the OCPD feeder. So a 600A circuit, feeding a panel protected by a 600A main, could have anything up to a full 600A of interconnected PV, which is 480A total current.

Case #2: There is an OCPD at both ends, but the feeder conductors aren't quite equal to it. Whomever sized the feeder, took credit for 240.4(B). In your 600A example, you would see this if the 600A feeder were built with 2 sets of 300 kcmil Cu (which add up to 570A).
Either A: With a load calculation to justify, decrease the load-side OCPD of the feeder, so it is less than the ampacity of the feeder conductors.
Or B: Replace the conductors from the tap point to the load-side OCPD, so they meet or exceed the downstream OCPD.
Or C: Replace the feeder conductors entirely, so they meet or exceed both OCPD's, making your situation be the same as Case #1.

For solution A and B to Case #2, you could then connect up to 500A worth of PV OCPD (i.e. 400A operating current). I.e. the standard OCPD rating, that is doesn't exceed the feeder's ampacity.

Case #3: There is NO OCPD at the load side, the subpanel is main lug only (MLO):
The feeder from the tap point up until the panel, must have an ampacity equaling PV OCPD + main supply OCPD. Same with the busbar of the MLO subpanel. So in your 600A example, if you desired to connect 200A of PV OCPD, you'd need and 800A rating of the MLO subpanel and 800A worth of conductors (such as 2x 600 kcmil Cu). The reason why this is an issue, is that the main supply + PV source could load that panel up to 800A, in the unlikely event that the user turns on more loads than the load calculation considers. One way to work around this, is to put a set of 600A fuses or separate 600A breaker, just before the MLO subpanel, and set it up to be like Case #1 or Case #2 instead.

#### henryo

##### Member
For this particular setup, you also have the 120% rule that applies on the 1200A panelboard. You apply the 120% rule on all panels from the point of interconnection, all the way up to the service point, which is why you usually want to interconnect on the main panel instead of subpanels. In NEC2014 and later, this would limit you to 80% of 240A, i.e. 192A. That would imply up to a 250A OCPD for the interconnection tap circuit, but the total operating current of the source would be limited to 80% of 240A instead of 80% of 250A.

If upstream equipment doesn't have its own rules that otherwise limit you, the rule for interconnecting on feeders in general is as follows. There are 3 cases, depending on whether an OCPD is at both ends of the feeder to protect from potential overload, and depending on whether 240.4(B) was applied when sizing the feeder. Typical of all 3 cases, the short section of conductor between the feeder and the first PV-specific OCPD would be constructed to comply with one of the 240.21(B) tap rules. Maintain symmetry when parallel conductors are involved.

Case 1: There is an OCPD at both ends of the feeder, and feeder conductors meet or exceed the OCPD protecting them (i.e. it sizing is not taking credit for 240.4(B)):
You can connect anything you can fit behind an OCPD, up to the full rating of the OCPD feeder. So a 600A circuit, feeding a panel protected by a 600A main, could have anything up to a full 600A of interconnected PV, which is 480A total current.

Case #2: There is an OCPD at both ends, but the feeder conductors aren't quite equal to it. Whomever sized the feeder, took credit for 240.4(B). In your 600A example, you would see this if the 600A feeder were built with 2 sets of 300 kcmil Cu (which add up to 570A).
Either A: With a load calculation to justify, decrease the load-side OCPD of the feeder, so it is less than the ampacity of the feeder conductors.
Or B: Replace the conductors from the tap point to the load-side OCPD, so they meet or exceed the downstream OCPD.
Or C: Replace the feeder conductors entirely, so they meet or exceed both OCPD's, making your situation be the same as Case #1.

For solution A and B to Case #2, you could then connect up to 500A worth of PV OCPD (i.e. 400A operating current). I.e. the standard OCPD rating, that is doesn't exceed the feeder's ampacity.

Case #3: There is NO OCPD at the load side, the subpanel is main lug only (MLO):
The feeder from the tap point up until the panel, must have an ampacity equaling PV OCPD + main supply OCPD. Same with the busbar of the MLO subpanel. So in your 600A example, if you desired to connect 200A of PV OCPD, you'd need and 800A rating of the MLO subpanel and 800A worth of conductors (such as 2x 600 kcmil Cu). The reason why this is an issue, is that the main supply + PV source could load that panel up to 800A, in the unlikely event that the user turns on more loads than the load calculation considers. One way to work around this, is to put a set of 600A fuses or separate 600A breaker, just before the MLO subpanel, and set it up to be like Case #1 or Case #2 instead.
Thank you. It is getting clearer and informative
Additional question on the 120% rule;
if both inverters A (for 600A Panel) and Inverter B (for 1200A panel) are installed at the same time ; is this a correct interpretation of the rule ...as long as Inverter A + Inverter B ---- complies to the 120% rule at the 1200A Panel or no greater than 192A per the example (see attached illustration) then it is ok.
Additional question on the Feeder tap:
With reference to your case 1 (OCPD at both ends; feeder are size no less than the OCPD; feeder tap rule applied) tapping inverter onto the feeder ...is it acceptable (say due to accessibility problem of the feeder) to do the tapping at the 600A panel incoming buss bar instead to the feeder wire itself (see illustration).
Additional comment: Among the three connection points in this specific scenario it is the feeder tapping that resulted to a greater inverter allowance.

#### Attachments

• 405.3 KB Views: 2

#### Carultch

##### Senior Member
Thank you. It is getting clearer and informative
Additional question on the 120% rule;
if both inverters A (for 600A Panel) and Inverter B (for 1200A panel) are installed at the same time ; is this a correct interpretation of the rule ...as long as Inverter A + Inverter B ---- complies to the 120% rule at the 1200A Panel or no greater than 192A per the example (see attached illustration) then it is ok.
Additional question on the Feeder tap:
With reference to your case 1 (OCPD at both ends; feeder are size no less than the OCPD; feeder tap rule applied) tapping inverter onto the feeder ...is it acceptable (say due to accessibility problem of the feeder) to do the tapping at the 600A panel incoming buss bar instead to the feeder wire itself (see illustration).
Additional comment: Among the three connection points in this specific scenario it is the feeder tapping that resulted to a greater inverter allowance.
For the first question, it is the sum total of inverters that is limited to 192A, whether you connect them at the 1200A panel, on the 600A feeder, or at a subpanel of the 1200A panel (assuming it is connected at the opposite end). Assume all local load might diminsh to zero, when applying it to a subpanel. Then carry the 120% rule all the way to the main panel.

I would not advise on connecting at multiple points of interconnection in multiple methods. While you may be able to put an interpretation of the rules together in theory, in practice you may end up confusing the rest of your team, or the AHJ when it comes time for inspection. Combine all sources first, and then interconnect. Aim for the main panel, and connect there. Rearrange the loads if necessary to get a space at the opposite end. If there is no space for your new breaker, you could remove a couple of existing loads, relocate to a new adjacent subpanel, and then you'd have an interconnection breaker in the main panel, and the breaker feeding that subpanel taking their place.

For the second question, and the left picture in your illustration, I think you misunderstood me. With the setup of the 1200A main panel, regardless of whether you connect on a feeder tap or opposite end branch breaker, or breaker on the subpanel, you'd be limited to 192A of what you can interconnect. You'd have to go to the line side of the 1200A main, in order to interconnect more. Where the 480A limit came from, was the hypothetical case that you had different service equipment, such that the 120% rule on the service equipment isn't the limiting factor. One example is a different setup of the main panelboard/switchboard, where the 120% rule has availability to backfeed it with the 600A source breaker. Another example is if you had a main switch that isn't part of a panelboard. Perhaps a 600A switch, that subsequently feeds a 600A panel with a 600A main. One reason you would see that, is if the utility requires cold-sequence metering.

When you construct a feeder tap, you can tap the feeder conductors themselves, or if you have luck, you could do it on spare lugs at equipment where they terminate. You might see 600A breakers built with 3 lugs per phase, and only 2 are used for the main 600A feeder. So there is an additional otherwise-unused lug, where you can make your tap. In any case, you'd have to do it on the line side of the 600A subpanel main breaker in your example. Once you get beyond that device, you no longer have the breaker protecting its panel against overload.

#### Carultch

##### Senior Member
To give you the big picture on all of this, remember that panelboards are routinely built with branch circuits that could potentially draw a lot more than the panelboard rating, when built for building loads. The load calculations depend on not all loads operating at their full capacity at once, in order that the panel is kept at a practical size. In the unlikely event that the loads do exceed the panelboard rating, we want the main breaker to trip, to prevent that from happening long enough to damage the panelboard.

If you have a second source, and the OCPD's aren't arranged to prevent it, there is a possibility that the loads could draw more than the main supply. This is why it is best that the sources be at opposite ends of the busbar. So that current cancels in the middle of the busbar, rather than accumulates as it would when fed from the same end. This is also why it is best that feeder taps be between breakers, rather than leading to an MLO panel.

#### henryo

##### Member
For the first question, it is the sum total of inverters that is limited to 192A, whether you connect them at the 1200A panel, on the 600A feeder, or at a subpanel of the 1200A panel (assuming it is connected at the opposite end). Assume all local load might diminsh to zero, when applying it to a subpanel. Then carry the 120% rule all the way to the main panel.

I would not advise on connecting at multiple points of interconnection in multiple methods. While you may be able to put an interpretation of the rules together in theory, in practice you may end up confusing the rest of your team, or the AHJ when it comes time for inspection. Combine all sources first, and then interconnect. Aim for the main panel, and connect there. Rearrange the loads if necessary to get a space at the opposite end. If there is no space for your new breaker, you could remove a couple of existing loads, relocate to a new adjacent subpanel, and then you'd have an interconnection breaker in the main panel, and the breaker feeding that subpanel taking their place.

For the second question, and the left picture in your illustration, I think you misunderstood me. With the setup of the 1200A main panel, regardless of whether you connect on a feeder tap or opposite end branch breaker, or breaker on the subpanel, you'd be limited to 192A of what you can interconnect. You'd have to go to the line side of the 1200A main, in order to interconnect more. Where the 480A limit came from, was the hypothetical case that you had different service equipment, such that the 120% rule on the service equipment isn't the limiting factor. One example is a different setup of the main panelboard/switchboard, where the 120% rule has availability to backfeed it with the 600A source breaker. Another example is if you had a main switch that isn't part of a panelboard. Perhaps a 600A switch, that subsequently feeds a 600A panel with a 600A main. One reason you would see that, is if the utility requires cold-sequence metering.

When you construct a feeder tap, you can tap the feeder conductors themselves, or if you have luck, you could do it on spare lugs at equipment where they terminate. You might see 600A breakers built with 3 lugs per phase, and only 2 are used for the main 600A feeder. So there is an additional otherwise-unused lug, where you can make your tap. In any case, you'd have to do it on the line side of the 600A subpanel main breaker in your example. Once you get beyond that device, you no longer have the breaker protecting its panel against overload.
I got it. Although the 600A feeder can handle 480A, the PV inverter size is limited to 192A dictated by the upstream panel rated at 1200A.
There is actually this commercial building (see illustration) whose 1200A Main panel is located (power house) quite far from the building itself and what it has is a 600A sub panel. The problem is that the building needs about 150A for its proposed PV Inverter and based on the calculation the 600A sub panel can only allow 96A. So applying this feeder tap the allowable capacity is 192A more that the 150A requirement. Looking at the feeder connection option (B1 & B2) I plan to connect directly o B2 as it seems to be more accessible and no more up sizing (600A + 125% of 150 additional requirement in the code) as the feeder is protected by a 600A OCPD itself.

#### Attachments

• 394.9 KB Views: 2

#### GoldDigger

##### Moderator
Staff member
This interpretation may well vary among AHJs and inspectors, but at least some will argue that the feeder needs to be treated the same as you would a panel bus. The justification is that you could at any point in the future connect an additional load to the same feeder between the PV connection point and the second panel.
As long as the feeder is sized so that it is properly protected by the upstream OCPD (600A) there are no restrictions on how much load you can add downstream on the same feeder. With the understanding that if you add too much and overload the upstream OCPD it will trip and the feeder will be protected.
But when you insert the PV source, and you then add an additional downstream load you now have the possibility that the total load will be more than 600A but the upstream OCPD will not trip because some of the load current is supplied by the upstream 600A breaker and some is supplied by the 96A PV. The result is that the feeder wire downstream of the PV connection point can be overloaded.
Other inspectors would allow this connection on the grounds that there is currently no other downstream load and there is no point in piling on future "ifs". In a panel, on the other hand, you have no expectation that there will not be other breakers and loads added later.
My take is that, whatever the local interpretation, it is a bad idea to make that tap.

#### jaggedben

##### Senior Member
This interpretation may well vary among AHJs and inspectors, but at least some will argue that the feeder needs to be treated the same as you would a panel bus. The justification is that you could at any point in the future connect an additional load to the same feeder between the PV connection point and the second panel.
...
That interpretation had a code-language basis in the 2011 NEC and earlier cycles, but I haven't heard it since our state went on the 2014 NEC equivalent, which changed the rules clearly and expansively to what Carultch has been describing above. Your mileage may vary. In particular, you might still encounter an AHJ who wants signage on the feeder connections to alert future electricians to applicable rules.