208Y/120V PANEL SCHEDULE KVA PER PHASE CALCULATION

[NICKWAIENG]

Hello,

I am an MEP electrical designer. For years I have always thought that if you have a 208V single phase load fed from a 208Y/120V panel, meaning there is a 2 pole breaker, and you want to know the per phase kVA to input into a panel schedule, you simply take the voltage (208v) and multiply by the FLA/MCA which in my case is 25 amps, then divide by 2 to get the kVA per pole for the panel schedule.

It would look like this:

208V * 25A = 5200VA = 5.2kVA

5.2kVA / 2 = 2.6kVA

So the first pole on the schedule would have 2.6kVA and the second pole would also have 2.6kVA.

I am receiving feedback on my schedule that the load per phase should acutally be calculated as:

5.2kVA/sqrt3 = 5.2kVA/1.73 = 3.002kVA, and the first pole on the schedule would have 3.002kVA and the second pole would also have3.002kVA.

Let me know what you guys think, Thanks!

 

[wwhitney]

It depends.

What is the actual current of just your 5.2 kVA load @ 208V? It's 5200/208 = 25A. What 120V load current would give you 25A? It's 25A*120V = 3.0 kVA, which matches the second method. While your first method gives a current of 2600VA / 120V = 21.7A.

So if you had a single 208V load as your only load on the panel, which method gives you the correct minimum panel size? The minimum panel size is 25A, so the second method is correct.

However, suppose you have a balanced set of three such 208V 2-wire loads, one A-B, one B-C, and one C-A. Then what is the line current from that balanced set? It's not 2 * 25A = 50A, it's 2 * 21.7A = 43.3A. [The ratio is sqrt(3)/2.] So for a panel whose only load is a balanced set of 208V loads, which method gives you the correct minimum panel size? It's the first method.

In other words, the first method is appropriate when you are dealing with a balanced set of 208V loads, while the second method is appropriate for any unbalanced portion.

Edit: way to combine the two methods in a subsequent post.

Note that both of these methods are conservative (when the first method is only used for balanced sets) approximations to the actual computation of line current. Which if you know the power factor angle of each load current would be determined by doing the full vector addition for each line based on the load currents connected to that line.

Cheers, Wayne

 

[wwhitney]

So suppose all your loads have the same power factor angle for simplicity, e.g. they are all resistive. Then you could do your load accounting like this:

For each line conductor, say A, keep track of 3 totals: "current from B" ; "current from N" ; and "current from C." 120V loads get assigned to "current from N", 208V A-B loads get assigned to "current from B", and 208V A-C loads get assigned to "current from C". Any 3 phase loads can be treated as a balanced set of L-N loads, so you can assign the line current to "current from N".

Then your single total starts off with just "current from N". To that, add the sqrt(3) times the minimum of "current from B" and "current from C". That represents how balanced currents add when they are 60 degrees apart. Subtract that minimum from "current from B" and "current from C", so you end up with one of those being zero. Add the remaining non-zero value to your total to get your final answer.

[This last step could be made more precise by recognizing that you are adding two currents that are still 30 degrees apart and doing the vector math, but the extra precision is likely not worth it.]

Cheers, Wayne

 

[electrofelon]

I have never understood the purpose of these panel schedules with loads. I can't think of any meaningful or useful information they would contain. It's not going to be the actual load and most likely won't even correspond to a NEC article 220 calc.

 

[winnie]

My summary: both approaches in the original post are approximations, both are strictly correct in (different) extreme cases, and both are likely close enough for any practical purpose.

If you need an exact answer, the only way to do it is with vector addition of all the load currents connected to each phase, as @wwhitney described.

IMHO the only reason for calculating the 'kVA per phase' is to document that you have a roughly balanced panel.

Jonathan

 

[electrofelon]

.

IMHO the only reason for calculating the 'kVA per phase' is to document that you have a roughly balanced panel.

Jonathan

I guess that makes sense. That will roughly balance the panel when the designer is not the installer, and the installer is an installer and not an electrician.