EV Charger loads in 208Y/120V Panel Schedule

[phila36]

Hi all. I recently received this panel schedule to review. I'm having trouble understanding how/why the total connected load came out to 72.7kVA.
For referene, the 20A/1P breaker is a 6A 120V load, and the 60A/2P breaker is a 48A 208V load.



According to my math:
(6A *120V) = 720VA
(48A * 208V) = 9,984 VA

(720VA + 9,984VA) * (6 chargers) = 64.2 kVA

This was an explanation I got:


I would greatly appreciate it if someone can tell me what i'm missing here.

Thank you!

 

[ggunn]

Are you missing that each charger only connects to two of the three phases?

 

[phila36]

Are you missing that each charger only connects to two of the three phases?

No, I understood that which is why I didn't apply the sqrt(3) to the VA calculations as these are all single phase loads.
This is a marked up schedule I sent back. Each charger only uses 2 poles but all three phases are still used for balancing:

 

[wwhitney]

According to my math:
(6A *120V) = 720VA
(48A * 208V) = 9,984 VA

(720VA + 9,984VA) * (6 chargers) = 64.2 kVA

Your math is correct, and because the load is perfectly balanced among the 3 phases, works for determining the highest phase current as well.

The explanation provided to you is correct, but as an explanation of why your math is good, and the schedule in the OP is wrong. Looking at phase A, when you add 48A from the EVSE on A-C to 48A from the EVSE on A-B, the sum is not 96A (which is what the panel schedule uses). Rather, the sum is 96 * (sqrt(3)/2) = 83A, because the current waveforms are 60 degrees out of phase.

This discrepancy of 13A, times 6 EVSES, times 120V = 9,360 VA is the principal difference between the schedule in the OP and your marked up schedule. The other difference is that you used 6A * 120V = 720 VA for the 20A circuit, while the schedule in the OP uses 5A.

Cheers, Wayne

 

[phila36]

Your math is correct, and because the load is perfectly balanced among the 3 phases, works for determining the highest phase current as well.

The explanation provided to you is correct, but as an explanation of why your math is good, and the schedule in the OP is wrong. Looking at phase A, when you add 48A from the EVSE on A-C to 48A from the EVSE on A-B, the sum is not 96A (which is what the panel schedule uses). Rather, the sum is 96 * (sqrt(3)/2) = 83A, because the current waveforms are 60 degrees out of phase.

This discrepancy of 13A, times 6 EVSES, times 120V = 9,360 VA is the principal difference between the schedule in the OP and your marked up schedule. The other difference is that you used 6A * 120V = 720 VA for the 20A circuit, while the schedule in the OP uses 5A.

Cheers, Wayne

on the second panel schedule with my red markups, isn't the out-of-phase difference you're referring to accounted for when I included the 1.73 when dividing the the 64,224VA by 208V?

 

[wwhitney]

on the second panel schedule with my red markups, isn't the out-of-phase difference you're referring to accounted for when I included the 1.73 when dividing the the 64,224VA by 208V?

Basically yes. Do you see something in my post as suggesting otherwise? I'm in agreement with your numbers, and was trying to explain where the other party went wrong, and to reconcile the two computations.

Cheers, Wayne

 

[ggunn]

...because the current waveforms are 60 degrees out of phase.

120 degrees, maybe?

 

[wwhitney]

120 degrees, maybe?

C -> A and A ->B are 120 degrees out of phase, but if you are adding currents on A, you flip one of those vectors. So you are adding two vectors 60 degrees out of phase, whence sqrt(3)/2.

Cheers, Wayne

 

[d0nut]

I truly do not understand why people do calculations like this in amps rather than in VA. In my opinion, it only adds to the chance of making errors and adds confusion because you have to be clear what voltage you are referencing as well when you give values in amps.

 

[ggunn]

I truly do not understand why people do calculations like this in amps rather than in VA. In my opinion, it only adds to the chance of making errors and adds confusion because you have to be clear what voltage you are referencing as well when you give values in amps.

For sizing conductors and OCPD, current is what counts.

 

[wwhitney]

For sizing conductors and OCPD, current is what counts.

Right, but if the load is fully balanced, then doing the calculation in VA and then dividing by V gives you an accurate answer for A.

If it's not fully balanced, you need to either use a larger fully balanced load (e.g. the feeder size for 5 EVSEs would be the same as for 6 EVSEs, assuming you want to keep your ungrounded conductors all the same size), or use vector math.

Cheers, Wayne

 

[phila36]

Basically yes. Do you see something in my post as suggesting otherwise? I'm in agreement with your numbers, and was trying to explain where the other party went wrong, and to reconcile the two computations.

Cheers, Wayne

Nope! Just still trying to work out the math to see how they got to 72.7 kVA connected load. By observation, I think what they did was 101A * 208V * 1.73 * 2 = 72,688 VA.
I'm not sure why they multiplied by 2 at the end there, but regardless, I see the mistake in adding up the amps and applying 208v to all loads (not all loads are 208v in the schedule and none of the loads are three phase.)

Can anyone figure why that VA value was multiplied by 2? Apologies if that was already covered, I'm still learning these principles.

 

[david luchini]

Can anyone figure why that VA value was multiplied by 2? Apologies if that was already covered, I'm still learning these principles.

Because 101A * 2 = 202A. They are showing a load current of 202A.

Or...202A * 208V * 1.73 = 72,688VA

 

[wwhitney]

Can anyone figure why that VA value was multiplied by 2? Apologies if that was already covered, I'm still learning these principles.

2 panels, each with 3 EVSEs.

While you can get to the VA from (the incorrect) 101A by going 101A * 208V * 1.73, I prefer to think of it as 101A * 120V * 3 (same answer). Because if all the loads were 120V L-N loads, then the panel schedule would be correct, and you'd do 101A * 120V + 101A * 120V + 101A * 120V (per panel).

Cheers, Wayne

 

[phila36]

2 panels, each with 3 EVSEs.

While you can get to the VA from (the incorrect) 101A by going 101A * 208V * 1.73, I prefer to think of it as 101A * 120V * 3 (same answer). Because if all the loads were 120V L-N loads, then the panel schedule would be correct, and you'd do 101A * 120V + 101A * 120V + 101A * 120V (per panel).

Cheers, Wayne

It's just one panel but I didn't account for the other side. Thank you for explaining!

 

[Jpflex]

No, I understood that which is why I didn't apply the sqrt(3) to the VA calculations as these are all single phase loads.
This is a marked up schedule I sent back. Each charger only uses 2 poles but all three phases are still used for balancing:

I don’t think you can call the system single phase just because you connect a load between a leg phase and a neutral.

Whether secondary output is delta or y it is still 3 phase so you’ll need to multiply voltage by square root of 3?

 

[wwhitney]

I don’t think you can call the system single phase just because you connect a load between a leg phase and a neutral.

The text you quoted doesn't call the system single phase. It said the loads are single phase, while the system is 3 phase. Any 2 wire load is single phase, as with just two conductors, there's only one voltage waveform, so there's no possibility of phase difference.

Whether secondary output is delta or y it is still 3 phase so you’ll need to multiply voltage by square root of 3?

On a wye, the L-L voltage is sqrt(3) times the L-N voltage. If your 2-wire load is connected L-N, you'd use that voltage. If it's connected L-L, you'd use the higher voltage with the factor of sqrt(3).

Cheers, Wayne

 

[Jpflex]

Any 2 wire load is single phase, as with just two conductors, there's only one voltage waveform, so there's no possibility of phase difference.

This doesn’t make sense because L1 and L2 are separate phases 120 degrees apart and if connected to a load I thought they were considered 3 phase

If a load were connected to 2 legs With phases 180 degrees apart then the load would be connected to a 2 phase system

 

[wwhitney]

This doesn’t make sense because L1 and L2 are separate phases 120 degrees apart and if connected to a load I thought they were considered 3 phase

Let me try putting it this way: to get a voltage waveform you need two wires. To get a phase difference, you need two different voltage waveforms. That means you need at least 3 wires.

In a 3 phase 4 wire wye system, you can say that L1-N and L2-N are separate phases and 120 degree apart. So terminologically we are apt to call L1 and L2 different phases. But that is only true relative to N (or L3). If you don't have N (or L3), and you just have L1 and L2, there's only one phase present.

In other words, say you open an unknown electrical box and see just two black wires inside (no neutral or ground for some reason) and measure 208V between them. Based on the voltage, you suspect that they are L1 and L2 (say) from a 3-phase system. But it's also possible they are from a 240V single phase system and a buck transformer, for some piece of equipment that needed 208V single phase. There's no way to tell the difference just measuring those two conductors at that point.

Cheers, Wayne

 

[LarryFine]

Any 2 wire load is single phase, as with just two conductors, there's only one voltage waveform, so there's no possibility of phase difference.

Correct.

This doesn’t make sense because L1 and L2 are separate phases 120 degrees apart and if connected to a load I thought they were considered 3 phase

No, the 120-vs-180-degree difference manifests itself by having 208v instead of 240v between the lines.

If a load were connected to 2 legs With phases 180 degrees apart then the load would be connected to a 2 phase system

Not. It's still one phase (see your own first sentence) with a polarity difference relative to the center tap.