Neutral Loads
- Thread starter Rahooty
- Start date
- Location
- Illinois
- Occupation
- retired electrician
I don't think those numbers are actually possible, so you can't prove it with math.
- Location
- New Jersey
- Occupation
- Journeyman Electrician
Why would you put the amp clamp around all three phases? Is there something that you're trying to read by doing so? Also the thread title is neutral loads are you trying to figure out the neutral current?
Carultch
Senior Member
- Location
- Massachusetts
To calculate neutral current, given phase currents, assuming you have the ideal case that there are no phase shifts, the formula is:
In = sqrt(Ia^2 + Ib^2 + Ic^2 - (Ia*Ib + Ib*Ic + Ic*Ia))
"Someone spilled SOS, in the house. In order to remove it, and have none left, we must SOP it up."
SOS = sum of squares, sounds like "sauce"
"In the house" = in the square root house
Remove = subtract
None starts with N, like neutral
SOP = sum of products
Applying this formula to your currents, I get In = 4.4 Amps.
When I set up the vector math, accounting for angles other than the ideal phase angles, I find that with phase shifts, I can achieve the neutral current. Assuming they both have the same phase shift, I get that when both phase shifts are 19.7 degrees, the neutral current can be 15.4A. This would be possible with power factors of 94%, of opposite types on these two phases, and unity power factor on phase A.
Here's the calculations:
| Current | Phase | x-direction (I * cos(phase)) | y-direction (I * sin(phase)) | ||
| Ia | 17.8 A | 0 deg | 17.80 A | 0 A | |
| Ib | 20.5 A | -139.74 deg | -15.64 A | -13.25 A | |
| Ic | 22.9 A | 139.74 deg | -17.48 A | 14.80 A | |
| Sum | -15.32 A | 1.55 A | |||
| Neutral current (negative of above) | 15.32 A | -1.55 A | |||
| Neutral current magnitude (Pyth combo of above) | 15.40 A |
Note that the phase angles for Ib and Ic should be -120 and +120 respectively, for the ideal case.
Clamping all three phases should read zero if they're in balance. I read 15 amps indicating an imbalance and I read the same 15 amps on the neutral.
I have a 3 phase distribution panel and I read the above current on each leg.
With the current I read in each leg I cannot figure out how the same meter comes up with 15 amps around the three phases and around neutral.
I have a 3 phase distribution panel and I read the above current on each leg.
With the current I read in each leg I cannot figure out how the same meter comes up with 15 amps around the three phases and around neutral.
Carultch
Senior Member
- Location
- Massachusetts
That's precisely what you should expect. The current among all phases and the neutral, has to add up to zero on a "healthy" circuit without any faults. All the current that departs the source, has to return to the source. This means that if you collectively ammeter any group of 3 conductors, the remaining conductor will have an equal and opposite current. So neutral should have current that is equal in magnitude and opposite in direction, from the collective current among the phase conductors.
wwhitney
Senior Member
- Location
- Berkeley, CA
- Occupation
- Retired
Notable that a power factor of 0.94, which sounds close to one, can cause Ineutral to increase in magnitude by approximately half of peak Iline.
Cheers, Wayne
Yes, so because I have a current other than zero, I have an imbalance, and I am not as good as you with my math lol, what I did is create a spreadsheet duplicating the distribution panel, it has 42 20A breakers and all but 3 or 4 are used.
To keep from disrupting the plant as little as possible, I am trying to swap current loads from phase to phase on the spreadsheet to find the best balance. But I'm having trouble matching my meter with my spreadsheet
Carultch
Senior Member
- Location
- Massachusetts
If every load has unity power factor, you simply add up the amps on all three phases, and try to distribute them to be as close as practical to equaling each other.
However, from the data you've provided, it appears you have non-unity power factor, which will cause some of the loads to not be at exactly 120 degrees from each other. This is where you'll need to account for power factor and phase angle, if you really want to eliminate neutral current.
I made a spreadsheet that allows you to input current and phase angle for each load, and add up the results. You can select the phase from a dropdown, and experiment with how it affects phase balancing and neutral current.
Phase Balancing w/ Phase Angle
Columns D & E are where you input your data, which currently contain randomly generated data. Enter negative phase angles for inductive loads, and positive phase angles for capacitative loads. You'll need to measure current and voltage simultaneously with a power analyzer to determine current and phase angle together. Phase A is at the 3 O'clock position on the +x axis, phase B is at the 7:00 position, and Phase C is at the 11:00 position.
Most of the time, if most of your loads are similar enough in ampacity, just laying out your loads randomly is "good enough", and this is all a moving target that is subject to user behavior. You can scientifically balance the loads to minimize neutral current, and you may only get marginally better. My recommendation would be to aim to balance your big ticket loads as closely as possible, use a few of your 20A and less loads to make up the shortfall where necessary, and then lay the rest of your 20A and less loads in any order among the rest of the positions.
Last edited:
Thank you Carultch for your much spent time, I saved your spreadsheet, and I am looking forward to studying it, but they have me in the annual mandatory MSHA training today.If every load has unity power factor, you simply add up the amps on all three phases, and try to distribute them to be as close as practical to equaling each other.
However, from the data you've provided, it appears you have non-unity power factor, which will cause some of the loads to not be at exactly 120 degrees from each other. This is where you'll need to account for power factor and phase angle, if you really want to eliminate neutral current.
I made a spreadsheet that allows you to input current and phase angle for each load, and add up the results. You can select the phase from a dropdown, and experiment with how it affects phase balancing and neutral current.
Phase Balancing w/ Phase Angle
docs.google.com
Columns D & E are where you input your data, which currently contain randomly generated data. Enter negative phase angles for inductive loads, and positive phase angles for capacitative loads. You'll need to measure current and voltage simultaneously with a power analyzer to determine current and phase angle together. Phase A is at the 3 O'clock position on the +x axis, phase B is at the 7:00 position, and Phase C is at the 11:00 position.
Most of the time, if most of your loads are similar enough in ampacity, just laying out your loads randomly is "good enough", and this is all a moving target that is subject to user behavior. You can scientifically balance the loads to minimize neutral current, and you may only get marginally better. My recommendation would be to aim to balance your big ticket loads as closely as possible, use a few of your 20A and less loads to make up the shortfall where necessary, and then lay the rest of your 20A and less loads in any order among the rest of the positions.
kwired
Electron manager
- Location
- NE Nebraska
- Occupation
- EC
Unless all are continuous steady loads I'd call it balanced enough as is in OP.