Panel Size with 120/208

[Davidsg]

Hello,

I have worked mostly with 120VAC panels, and this is the first time that I will be working with 120 and 208 VAC in the same panel.

The devices that I need to work with are 5 chargers, a controller and 3 workstations.

The chargers are going to be fed with 208VAC, they have a max current of 15A (L-L) each.

The controller is going to be fed with 120VAC, with a max current of 8A (L-N).

The workstations are going to be fed with 120VAC, with a max current of 4A (L-N) each.

As for the chargers, that are going to be fed with 208VAC L-L, the current has to be taken into account for each line ? Below I left the panel distribution:

	A	B	C
WS1	4
WS2		4
WS3			4
CONTROLLER	8
CHARGER 1		15	15
CHARGER 2	15	15
CHARGER 3	15		15
CHARGER 4		15	15
CHARGER 5	15	15
TOTAL	57	64	49

If this is true, then the FLA of this distribution would be 167A, so a 200A panel would be good to go ?

Thank you in advance.

 

[infinity]

If this is true, then the FLA of this distribution would be 167A, so a 200A panel would be good to go ?

You do not add the A, B, and C phase currents together. The maximum load on any one phase according to your table is 64 amps. Maybe an engineer can explain the calculation using VA.

Welcome to the Forum.

 

[winnie]

The rating convention is that circuits are rated by the trip current on a single phase or the ampacity of a single conductor. So a 200A circuit would have 200A of capacity on each of the hot wires, not some sort of total that equates to 200A.

Similarly, a 200A 120V single phase panel can provide 200A on 1 hot bus (24000 VA) , a 200A 120/240V split phase panel can provide 200A on each of its two hot busses (48000 VA), and a 200A 208/120V three phase panel has _three_ busses, each with 200A capacity (72000 VA). This is just a naming convention, not physics; the manufacturers could have equally decided that all '200A' panels would have the same capacity.

The calculation that you did is approximate but good enough for sizing a panel.

A similar approximation is the VA approach:
1) Calculate the VA for each load. For single phase loads this is load current * load voltage. For 3 phase use load current * load voltage * 1.732
2) Apportion the VA to each connected phase:
a) A 120V load on a single phase, the whole VA goes to the phase
b) A 208V single phase load, half of the load VA goes to each connected phase.
c) A 208V three phase load, 1/3 of the load VA goes to each connected phase.
3) Sum up each phase VA
4) Each phase VA / 120V is the approximate phase current.

-Jonathan

 

[HarrySlother]

Yeah, unless all the loads are 3 phase you need to do it the long way and calculate the VA on each phase

 

[wwhitney]

The calculation that you did is approximate but good enough for sizing a panel. A similar approximation is the VA approach:

For the case that there are no 208V loads, both methods are equivalent and exact.

Then with 208V loads, if one line has both a 208V load and a 120V or 3 phase load, then the "amps" method is an overestimate. For the VA method, if the 208V loads are fully balanced, it is exact. Whereas if the 208V loads are not fully balanced, the VA method is an underestimate. (E.g. look at the case of a single load that is 208V.)

So a conservative approach that doesn't penalize you for balanced sets of 208V loads is to first make as many groups of balanced 208V loads as you can, and convert those into the equivalent 3 phase load (e.g. via the VA method). Then use the amps method. For the OP, that gives values that are 15*(2-sqrt(3)) = 4 A less on each line than the calculation in the OP. Which is still more than the nonconservative result in post #4.

Cheers, Wayne

 

[Davidsg]

Hello everyone

Thank you for the replies and warm welcoming.

So this is the summary of what you have explained to me :

The rating convention is that circuits are rated by the trip current on a single phase or the ampacity of a single conductor. So a 200A circuit would have 200A of capacity on each of the hot wires, not some sort of total that equates to 200A.

Similarly, a 200A 120V single phase panel can provide 200A on 1 hot bus (24000 VA) , a 200A 120/240V split phase panel can provide 200A on each of its two hot busses (48000 VA), and a 200A 208/120V three phase panel has _three_ busses, each with 200A capacity (72000 VA). This is just a naming convention, not physics; the manufacturers could have equally decided that all '200A' panels would have the same capacity.

So the panel rating is for each bus ? So if I get a 200A panel, each hot can hold up to 200 A ?

The calculation that you did is approximate but good enough for sizing a panel.

A similar approximation is the VA approach:
1) Calculate the VA for each load. For single phase loads this is load current * load voltage. For 3 phase use load current * load voltage * 1.732
2) Apportion the VA to each connected phase:
a) A 120V load on a single phase, the whole VA goes to the phase
b) A 208V single phase load, half of the load VA goes to each connected phase.
c) A 208V three phase load, 1/3 of the load VA goes to each connected phase.
3) Sum up each phase VA
4) Each phase VA / 120V is the approximate phase current.

With the explanation, the VA method is good to follow. Also, didn't know that the current was going to be splitted up, as I thought that as we are using two phases, I would think that the phases are connected in series so the current flowing through them should be the same on both.

View attachment 2578107

So in this case, the panel rating should be 75A instead of 175-200A ?

Thank you in advance.

 

[wwhitney]

So the panel rating is for each bus ? So if I get a 200A panel, each hot can hold up to 200 A ?

Yes.

With the explanation, the VA method is good to follow.

But still sometimes nonconservative as I described. Your original method is conservative.

Also, didn't know that the current was going to be splitted up, as I thought that as we are using two phases, I would think that the phases are connected in series so the current flowing through them should be the same on both.

The current isn't really getting split up, it's the apparent power that is split up, which for L-L loads is 208V * the current. And then when the apparent power is changed back into current, you use 120V. So the current ends up getting multiplied by 1/2 * 208 / 120 = sqrt(3)/2. Which is correct only when all the L-L loads are in balanced sets of 3.

Cheers, Wayne

 

[winnie]

Hello everyone

So the panel rating is for each bus ? So if I get a 200A panel, each hot can hold up to 200 A ?

Yes.

With the explanation, the VA method is good to follow. Also, didn't know that the current was going to be splitted up, as I thought that as we are using two phases, I would think that the phases are connected in series so the current flowing through them should be the same on both.

The _current_ is the same on both connected legs, because it is a series circuit. The VA 'assigned' to each leg in the calculation is split up. As @wwhitney explained, this comes out in the wash because the VA for the single phase load is at 208V but the VA per leg is calculated at 120V.

So in this case, the panel rating should be 75A instead of 175-200A ?

75A would be more than sufficient. You can probably go smaller, but the effort of figuring out the _exact_ amount smaller is not worth the savings in materials; you will probably use a 100A panel because they are hard to find smaller, so the sizing question is more for the circuit feeding the panel.

-Jonathan

 

[wwhitney]

The _current_ is the same on both connected legs, because it is a series circuit. The VA 'assigned' to each leg in the calculation is split up. As @wwhitney explained, this comes out in the wash because the VA for the single phase load is at 208V but the VA per leg is calculated at 120V.

It comes out as a factor of 0.866, which is correct for balanced sets of L-L loads only.

Cheers, Wayne

 

[winnie]

It comes out as a factor of 0.866, which is correct for balanced sets of L-L loads only.

Cheers, Wayne

Absolutely, but this is a rough approximation to see if the panel is big enough.

When the load is rated 15A at 208V it almost certainly draws less than this, and the loads will have diversity that we can't quantify, and we are getting a panel that is bigger than the number we come up with. At this level of approximation 208/120 = 2

-Jonathan

 

[wwhitney]

Absolutely, but this is a rough approximation to see if the panel is big enough.

But it's a non-conservative approximation. The method in the OP is a conservative approximation, and simpler.

Cheers, Wayne

 

[winnie]

Fair enough. The VA calculation helps understand the original calculation. But you only need to do one, and the original calculation is more conservative

 

[wwhitney]

So I'd say that the upshot is that the VA method is only useful as follows: if the simple "add up the line currents as scalars" method is slightly over a given panel size, try the VA method. If that's still over the given panel size, you know you need to upsize the panel. But if it's under, you need to do a more careful accounting to better approximate the actual vector math. Or just do the actual vector math.

Cheers, Wayne