Unbalanced 3 Phase Panel Calculations

[thesink79]

To preface this, I am totally new to the electrical world. I graduated college last summer and started as a drafter/designer with an electrical contractor. We use an excel based calculator to general panel schedules and their calculations. When it accounts for the base load of the panel instead of taking a sum of the VA on each phase it multiplies the largest phase by three. In panels that are well balanced there is not much of a difference. However in panels that are not balanced and cannot be balanced because of large loads this can create a large difference in overall VA and then AMPS. Here is an example of a panel that is balanced the best I could get it but still not balanced:

Line 1: 6,218 VA
Line 2: 8,320 VA
Line 3: 4,990 VA

Largest Line Load X 3: 24,960

Sum of Lines: 19,524 VA that's a 22% difference on top of a 20% factor of safety applied later in the calculator, it seems overcalculated for unbalanced systems.

I am trying to find out where in the code or the reasoning behind taking the largest load x 3.

I get that its basically auto balancing on the largest load but where in the code does it say we need to do that or what the practical reason other than a safety factor?

I asked the owner of the calculator and his answer for me was "The ampacity of the largest load on a single conductor determines the conductor size per NEC requirements". Is he saying we determine our overall load on a panel based on the conductor size and that the conductor size is based on the largest line load? Is that correct?

Thanks for the help.

 

[wwhitney]

. . .

Largest Line Load X 3: 24,960

Sum of Lines: 19,524 VA that's a 22% difference

. . .

I asked the owner of the calculator and his answer for me was "The ampacity of the largest load on a single conductor determines the conductor size per NEC requirements". Is he saying we determine our overall load on a panel based on the conductor size and that the conductor size is based on the largest line load? Is that correct?

The load on the panel is still 19,524 VA. But panels come with equally sized busses on each line, and feeders generally have equally sized conductors for all the ungrounded conductors. So both the bus size and the feeder conductor size is determined by the worst case load on a single line, which is Line 2 at 8,320 VA in your example.

In other words, you have to size everything as if Line1, Line2, and Line3 were all 8,320 VA. But if you ever needed to add additional load, you'd have some reserve capacity on Lines 1 and 3 for L-N or L-L loads--you could increase those VA up to 8,320 VA without changing your worst case load on a single line.

BTW, this VA method is just a convenient accounting approximation; ultimately what matters is the actual amps on each line conductor or panel bus. If you have a mixture of L-N, L-L, and L-L-L (3 phase) loads of known power factor, there are more exact methods for determining the amps on each line than the VA method. Generally not worth the extra computational complexity.

Cheers, Wayne

 

[electrofelon]

The load on the panel is still 19,524 VA. But panels come with equally sized busses on each line, and feeders generally have equally sized conductors for all the ungrounded conductors. So both the bus size and the feeder conductor size is determined by the worst case load on a single line, which is Line 2 at 8,320 VA in your example.

In other words, you have to size everything as if Line1, Line2, and Line3 were all 8,320 VA. But if you ever needed to add additional load, you'd have some reserve capacity on Lines 1 and 3 for L-N or L-L loads--you could increase those VA up to 8,320 VA without changing your worst case load on a single line.

BTW, this VA method is just a convenient accounting approximation; ultimately what matters is the actual amps on each line conductor or panel bus. If you have a mixture of L-N, L-L, and L-L-L (3 phase) loads of known power factor, there are more exact methods for determining the amps on each line than the VA method. Generally not worth the extra computational complexity.

Cheers, Wayne

I believe in theory you could have a feeder with different sized conductors (each protected with the appropriate size OCPD) feeding a panelboard, right?

 

[wwhitney]

I believe in theory you could have a feeder with different sized conductors (each protected with the appropriate size OCPD) feeding a panelboard, right?

I guess in theory. But 240.15(B) often requires that all ungrounded conductors be opened simultaneously. Can you get 3 pole OCPD with unequal ratings on the 3 poles?

It's unclear to me whether 240.15(B)(3), which covers "line-to-line" loads on 208Y/120V systems and allows handle-tied single pole breakers, is meant to cover "L-L-L" loads like a feeder.

Cheers, Wayne

 

[electrofelon]

I guess in theory. But 240.15(B) often requires that all ungrounded conductors be opened simultaneously. Can you get 3 pole OCPD with unequal ratings on the 3 poles?

It's unclear to me whether 240.15(B)(3), which covers "line-to-line" loads on 208Y/120V systems and allows handle-tied single pole breakers, is meant to cover "L-L-L" loads like a feeder.

Cheers, Wayne

I was thinking a fused safety switch to feed this abomination

 

[ptonsparky]

I was thinking a fused safety switch to feed this abomination

You would see Hi leg Delta services with this arrangement often enough.

 

[Strathead]

I can't imagine that a system couldn't be redesigned to greatly reduce that imbalance. I can understand an existing building that has been Frankensteined but, The only scenario I can see that would cause the above would be one huge single phase load. Why does the OP say it can't be balanced better?

 

[Carultch]

I am trying to find out where in the code or the reasoning behind taking the largest load x 3.

I get that its basically auto balancing on the largest load but where in the code does it say we need to do that or what the practical reason other than a safety factor?

I asked the owner of the calculator and his answer for me was "The ampacity of the largest load on a single conductor determines the conductor size per NEC requirements". Is he saying we determine our overall load on a panel based on the conductor size and that the conductor size is based on the largest line load? Is that correct?

Thanks for the help.

I think it is more of a KISS principle of design that the author of that spreadsheet gave this recommendation. As in, "keep it simple stupid". By specifying the same size for all phase conductors, you mitigate the risk that the installer misinterprets the intents of your plans.

Specifying wildly different phase conductor sizes, just because you can, might result in a setup that no longer complies if someone makes a mistake on following your diverse design. In theory, you can use a fused disconnect with different fuses to protect different sized conductors. I could see that happening for a high leg system where the B phase takes significantly less load. But for a WYE system, I don't see the point of the extra complexity. You're better off putting in the effort to balance the phases in the first place.

Your wires ultimately do need to be protected by the OCPD, and since that may be a breaker, you'll inevitably have the same trip rating on all three poles.