3phase 240/120 High Leg Delta

Status
Not open for further replies.
I thought I understood 3-phase power perfectly well until I had to size a panel with 240 3-phase High Leg Delta service. Going through old forum posts is getting me more confused.

To get the current on each bus bar (line current), a panel schedule I have adds up all the VA on each phase and divides them by 120V for bus bars A and C, and divides by 208V for the "high leg" bus bar B.

Does this sound correct to you all?

(Please forgive me- I'm used to working with 208Y/120V 3 phase)

The way I envisioned doing this was different: I thought to use separate calculations for 3-phase balanced loads, 2-pole single phase loads, and 1-pole single phase loads.

Say my 3-phase loads add up to 10,000VA. Recalling that Total_VA = Vline_line*I_line*1.732, I calculated 10,000/(240V*1.732)=24A in each bus bar

Right off the bat this differs from the panel schedule macro, which gives me 27.8A on legs A and C, and 16A on the high leg B.

I'll leave it at that for now. Thanks in advance for any help!
 
Last edited:

GoldDigger

Moderator
Staff member
Location
Placerville, CA, USA
Occupation
Retired PV System Designer
The only problem that I see with that general approach, without looking in detail at your numbers, is that whenever you have two line to line loads sharing a common phase line, just adding the currents for them will overstate the line current, since the current to the two loads will not be in phase. Instead there will be a 120 degree phase difference.
Your calculation is conservative, in that no actual line current will excede that value, even with any arbitrary power factor phase angles among the loads.
 

winnie

Senior Member
Location
Springfield, MA, USA
Occupation
Electric motor research
Your first approach (calculating for balanced 3 phase loads, separately calculating for single phase loads and then adding things up) will be a pretty good approximation to the actual load.

Balanced 3 phase load current is supposed to be the same on all three legs.

Dividing by 208 for the 'high leg' only makes sense if the loads on that leg were single phase 'line to neutral' loads, which they should not be :).

As GoldDigger mentions, the 'three phase' portion of the load on phase A will be out of phase with the 'single phase' portion, so it is not strictly correct to add these numbers up, but it will be close enough and a slightly high (conservative) estimate. (The phase angle difference will be about 30 degrees.)

-Jon
 

Smart $

Esteemed Member
Location
Ohio
I'd enter line VA as normal... one-third for 3 phase loads, one-half for line-to-line single phase, and full VA for line to neutral, then divide column VA totals by 138.6V [i.e. 240V/sqrt(3)].
 
The only problem that I see with that general approach, without looking in detail at your numbers, is that whenever you have two line to line loads sharing a common phase line, just adding the currents for them will overstate the line current, since the current to the two loads will not be in phase. Instead there will be a 120 degree phase difference.
Your calculation is conservative, in that no actual line current will excede that value, even with any arbitrary power factor phase angles among the loads.

Thank you for chiming in.

I'm most curious in being exactly correct, as now I'm second guessing my methods (even with straightforward 208Y/120V calcs).

[keep in mind this is all calculating the current going through my panel bus bars, not the transformer coils]

For simplicity, let's say my entire building loads are as follows (assume these already include any demand factors/continuous load adjustment, etc):
-A chiller, 240V 3 phase, 15000VA, on A, B, C
-A dishwasher, 240V 2 pole single phase, 3000VA, on A, B
-A condenser unit, 240V 2 pole single phase, 4000VA, on A, C
-A 240V receptacle load, 240V 2 pole single phase, 1000VA, on B, C
-Lighting, 120V 1 pole, 1000VA, on A, N
-120V Receptacles, 120V, 1000VA, 1 pole on C, N

If you'll notice, I strategically included 1 of each type of load: 3 phase A-B-C, 2pole A-B, 2pole A-C, 2pole B-C, 1pole A-N, and 1pole C-N

Chiller, 3 phase, A, B, C:
15,000/(240*1.732) = 36A on bus bars A, B, C
or 5000VA on each phase, so (5000/240)*1.732 = 5000/138V = 36A on bus bars A, B, C

Dishwasher, 2 pole, A, B:
1500VA/240V = 6.25A on bus bar A and bus bar B

Condenser, 2 pole, A, C:
2000VA/240V = 8.33A on bus bar A and bus bar C

240V Receptacle, 2 pole, B, C:
500VA/240V = 2.1A on bus bar B and bus bar C

Lighting, 1 pole, A, N:
1000VA/120V = 8.33A on bus bar A

120V receptacles, 1 pole, C, N:
1000VA/120V = 8.33A on bus bar C

So now I add them up:
A = 36 + 6.25 + 8.33 + 8.33 = 58.91
B = 36 + 6.25 + 2.1 = 44.35
C = 36 + 8.33 + 2.1 + 8.33 = 54.76

Now, my macro panel schedule excel sheet does it this way:
Bus Bar A: 5000 (chiller) + 1500 (dishwasher) + 2000 (condenser) + 1000 (lighting) = 9500VA
9500VA/120V = 79.17A

Bus Bar B: 5000 (chiller) + 1500 (dishwasher) + 500 (240V receptacle) = 7000VA
7000VA/208V = 33.65A

Bus Bar C: 5000 (chiller) + 2000 (condenser) + 500 (240V receptacle) + 1000 (120V receptacles) = 8500VA
8500VA/120 = 70.83A


There are not at all similar. Who's method is more accurate?
 
winnie: Dividing by 208 for the 'high leg' only makes sense if the loads on that leg were single phase 'line to neutral' loads, which they should not be :).

Thank you, I agree. That was driving me crazy.

Smart $: I'd enter line VA as normal... one-third for 3 phase loads, one-half for line-to-line single phase, and full VA for line to neutral, then divide column VA totals by 138.6V [i.e. 240V/sqrt(3)].


Ah yes- that was my first approach, which will give me yet another unique set of currents! Is it 'most' correct to do it this way? (Divide each VA total by 138V)

A: 9500/138.6V = 68.5A
B: 7000/138.6V = 50.5A
C: 8500/138.6V = 61.3A

Sorry for being so tenacious- it just really bugs me.
 
Last edited:

Smart $

Esteemed Member
Location
Ohio
...
Ah yes- that was my first approach, which will give me yet another unique set of currents! Is it 'most' correct to do it this way? (Divide each VA total by 138V)...
It is the most correct panel schedule method short of including power factor (phase angles) and doing the vector math.
 

kwired

Electron manager
Location
NE Nebraska
Most of the time when you have this particular system you either have:

1: limited three phase load and a lot of single phase load - say a commercial building with a three phase AC being the only three phase load and the rest of the load is essentially 120/240 single phase. - you likely figure each separately and then add the two "single phase" ungrounded lines together and the high leg just kind of follows along though it is probably not really loaded all that much, you are essentially calculating a single phase service for the most part.

2: majority of load is three phase motors, and you maybe have some limited 120 volt loads for lighting or other general purpose uses.
here you likely figure the three phase load and depending on how significant the single phase load is, you may just figure it as balanced across all phases or if somewhat significant you probably do figure it separately and add to the two "single phase ungrounded lines" and again the high leg just kind of follows along though it doesn't get anything added for this portion.

3: Open delta systems - in particular in remote areas where maybe POCO saves some cost by only running two phase and one neutral for the primary. This instance can have loads from either situation 1 or 2 and you probably still do load calcs according to 1 or 2, but POCO possibly needs to derate transformers used depending on how it is to be loaded.
 
Status
Not open for further replies.
Top