Balancing single phase loads from 3 phase

Status
Not open for further replies.

GITRDUN

Member
Location
Ks
First post so go easy on me, and yes im electrically challenged.

I have a large CNC machine shop powered by a 480v Tx feeding a 208v Wye Tx. Our single phase power is connected to the 208v. Awhile back we checked the loads being drawn on all 3 phases coming from our 208 and noticed that they are grossly unequal, one leg has about double the load as the other two. I called the electrician to ask if this is a problem and he replied - not at all, its normal because the single phase is all drawn from one leg. We are experiencing a lot of problems with our machines electronics boards and i highly suspect its a ground or neutral problem but as im no electrician i really dont know and thats a VERY long discussion for another day.

So i am curious if it is sound practice to split the single phase load equally from all three legs rather than run all single phase from one leg. Or is that even possible? From what i gather unbalanced loads on 3 phase is highly inefficient.
 
I allowed this post as the OP is not asking any DIY question.
 
Ask your electrician to switch the single phase loads to roughly balance the single phase loads among the phases.

Poor practice to load all onto one phase.

re: problems with our machines electronics boards --
If there are a lot of single phase loads on one leg that are also electronics, many power supplies of computers and similar are simply a full wave rectifier bridge, which produce a LOT of harmonics.

I have a Tree 325 CNC mill at home, and know that many of the electronic circuits in that mill operate off single phase. A high harmonic content on some of those circuits can lead to internal overvoltages and shutdown; had one case we investigated at a different shop where the harmonics were high enough to cause an LC resonance of an EMI filter and blow out the EMI filter cap due to overvoltage casued by the resonance.

Additionally, a heavily load single phase can pull the voltage down of that phase and also cause a phase shift due to the reactive impedance of the transformer. This results in high negative sequence components which will reduce non-vfd drive 3 phase motor torque.
 
quite a load

quite a load

Might I also suggest calling a different electrician. He should have told you that balanced loads are better, even if he didn't understand the impact of harmonics from the power supplies.

Bob
 
Balancing the load is desirable, but not always attainable. You should try to balance the larger more continuous loads, but if you have a shop full of fairly intermittent loads, complete balancing will likely never happen.

This usually becomes even more important when running from on site power sources, as those sources tend to have more direct negative results from imbalance then a utility source.
 
We just measured the amp laod on all 3 phases and neutral coming from our 208Y Tx. Blue leg has 65A, Middle leg 47A, Red leg 73A, and neutral had 31A. The amp load rises and falls continously as machines start and stop motors and such but we were able to capture the amp load on all phases while things were not fluctuating up and down. Voltage across all 3 phases measures 120V from phase to ground and phase to neutral. Measuring neutral to an unconnected ground rod 10ft away gave .03v, maybe a useless measurement, i dont know. The neutral from the 208Y Tx is connected to a ground bar in the main box where there is a ground wire connected running to a ground rod outside the building and all ground connections for the building are connected at that point in the box.

So is this as bad as i was thinking or is this not that bad after all. Also what about the 31A on the neutral leg? Is that ok? Is the amp load on the neutral increasing my electric bill or does that current just get recycled back at the Tx?
 
...

So is this as bad as i was thinking or is this not that bad after all. Also what about the 31A on the neutral leg? Is that ok? Is the amp load on the neutral increasing my electric bill or does that current just get recycled back at the Tx?
Hard to say. Definitely not ideal at the instances you took the measurements. As you noted, the current changes with loads getting turned on and off. Just using a one-reading meter, the powered loads could have changed in the time it takes to go from one conductor to another to get the current measurement. The only way to evaluate with any certainty is to connect a recording monitor which captures all 4 conductors current over time.

Balancing of a system is usually from the perspective of all loads being on at the same time... and in most cases does not take into consideration which ones will be on or off at different times.

The 31A on the neutral doesn't seem abnormal. No way to tell without knowing what loads were on when measured. Your electric meter measures net power transferred. Makes no difference which conductors carry the current.
 
Also what about the 31A on the neutral leg

If your phases were exactly 120 deg separation at unity power factor, you should see 24A neutral. (If did that calc right quick)

Which mean that probably phase rotation could be offset and there are fairly low power factors on one or all 3 phases and/or there are appreciable harmonics.


As an old mentor said, "ya prolly gots high harmonics if the white wire has turned brown"

To really diagnose problems, you need a 3 channel scope to look at all 3 phase simultaneously.
 
Also what about the 31A on the neutral leg

If your phases were exactly 120 deg separation at unity power factor, you should see 24A neutral. (If did that calc right quick)

Which mean that probably phase rotation could be offset and there are fairly low power factors on one or all 3 phases and/or there are appreciable harmonics.


As an old mentor said, "ya prolly gots high harmonics if the white wire has turned brown"

To really diagnose problems, you need a 3 channel scope to look at all 3 phase simultaneously.

The fact that his phase readings are changing as he take them would cause me not to be to concerned if the "math" didn't work since by the time he read the 31 amps on the neutral, the phase was probably different from when it was read. Just my opinion.
 
Good point, that is why the recommendation to take all 3 phase reading simultaneously with a scope.
 
If there are a lot of single phase loads on one leg that are also electronics, many power supplies of computers and similar are simply a full wave rectifier bridge, which produce a LOT of harmonics.

I have a Tree 325 CNC mill at home, and know that many of the electronic circuits in that mill operate off single phase. A high harmonic content on some of those circuits can lead to internal overvoltages and shutdown;

The harmonics thing has me wondering. I have also noticed that although our CNC machines are 3 phase it always seems that one leg draws much more current than the other two when taking measurements on the wires going directly into a particular machine. I assumed that was from the single phase usage for most of the machines components. Even when running the spindle motors if i remember correctly it seemed to pull mostly from one leg (been awhile since i checked all that but that seems the way i recollect it). So if all of our machines are running their single phase components from the same leg then i would assume this can possibly create alot of the harmonics you all are talking about. If that is true then does anyone know if it is possible for me to have my electrician switch the 3 phase wires around on each machine to spread the single phase usage around to all 3 legs instead of the same leg being used for single phase on all machines? We have always put the middle leg in the middle on the machines and red and blue were whichever side they need to be on to get the motors turning in correct direction. So im guessing you can not put the middle leg anywhere but in the middle, or can you?

Thanks for the help fellas. I know just enough about electrics to know i have no business doing things on my own. I will certainly have my electrician make any changes that need to be made but i need to be able to explain to him exactly what it is i want changed.
 
Unless one leg is grounded or is a "high-leg" on a 120/240 three phase four wire system, the NEC does not care which wires you put where when connecting to your load equipment.
 
The harmonics thing has me wondering. I have also noticed that although our CNC machines are 3 phase it always seems that one leg draws much more current than the other two when taking measurements on the wires going directly into a particular machine. I assumed that was from the single phase usage for most of the machines components. Even when running the spindle motors if i remember correctly it seemed to pull mostly from one leg (been awhile since i checked all that but that seems the way i recollect it). So if all of our machines are running their single phase components from the same leg then i would assume this can possibly create alot of the harmonics you all are talking about. If that is true then does anyone know if it is possible for me to have my electrician switch the 3 phase wires around on each machine to spread the single phase usage around to all 3 legs instead of the same leg being used for single phase on all machines? We have always put the middle leg in the middle on the machines and red and blue were whichever side they need to be on to get the motors turning in correct direction. So im guessing you can not put the middle leg anywhere but in the middle, or can you?

Thanks for the help fellas. I know just enough about electrics to know i have no business doing things on my own. I will certainly have my electrician make any changes that need to be made but i need to be able to explain to him exactly what it is i want changed.
A couple or three points in no particular order.

If as you indicate, one leg draws much more current at the CNC machine itself, you probably don't want the electrician messing about with the CNC internal wiring.
If it's on the common supply then you can probably move the phases round but will possibly need to do it in such a way that the phase rotation remains the same. Your electrician should know how to do that.

Harmonics is sometimes seen as a bit of a black art and is sometimes blamed as a catch all for anything even if it has more prosaic explanations that just haven't been investigated.
The computer part of the CNC will take non-linear currents as will anything else with a bit of control electronics and non-linear equates to harmonics. Even if that has a high percentage of non-linear, it probably isn't a lot in absolute terms.

So what else? The moving parts. The spindle motor, if speed controlled, will generate harmonics. If those controllers are single phase,third harmonic can be an issue particularly if you have a bunch of CNCs being fed from a common supply. This can be a sticky issue. You can, as suggested above, move the phases round the three phases to balance the common supply current in the three phases but that doesn't eliminate the harmonic currents. There is no simple answer other than ensuring that the supply conductors are not overloaded.
 
Unless one leg is grounded or is a "high-leg" on a 120/240 three phase four wire system, the NEC does not care which wires you put where when connecting to your load equipment.

Bingo. I had forgotten the fact that we did have a 240 high leg delta setup before it was changed over to a 208Y.:slaphead: This explains why all of our single phase is ran from the same leg and all of our 3 phase machines have the middle leg in the middle.

Also i was mistaken about the machines spindle motors pulling from only one leg. We checked a few of them yesterday and the motors do pull from all 3 legs pretty evenly.
 
I had my maint man work a little on the electrics today. He balanced the single phase loads a little more evenly. No way to get it all perfectly even but it is better and the current on the neutral is half of what it was. I dont understand how that is but then again i dont know much about this stuff. Also our single phase loads were distributed aross both outside legs rather than one single leg as i had thought.

One thing that really stuck out to me is the rating on our 208Y transformer. It is rated at 225KVA 3.5% Imp. Its a 480 to 208Y. Adding up the amp loads from all 3 phases plus the neutral came to about 72 amps. Thats with about 80% of our single phase loads running and no 3 phase machines running. The machines pull around 15amps each and usually have about 5 machines running from this transformer at one time. So that puts us around 150amps theoretically. Unless i am not figuring correctly (which is highly likely) i figure that we are only using about 15% of the transformers capacity.

If this is true would it be more economical to replace this Tx with a smaller one, or possibly put a small Tx at each machine and one larger one for the single phase? Im curious if we are burning a ton of electricity just by having this big transformer on whether its loaded or not.
 
I had my maint man work a little on the electrics today. He balanced the single phase loads a little more evenly. No way to get it all perfectly even but it is better and the current on the neutral is half of what it was. I dont understand how that is but then again i dont know much about this stuff. Also our single phase loads were distributed aross both outside legs rather than one single leg as i had thought.

One thing that really stuck out to me is the rating on our 208Y transformer. It is rated at 225KVA 3.5% Imp. Its a 480 to 208Y. Adding up the amp loads from all 3 phases plus the neutral came to about 72 amps. Thats with about 80% of our single phase loads running and no 3 phase machines running. The machines pull around 15amps each and usually have about 5 machines running from this transformer at one time. So that puts us around 150amps theoretically. Unless i am not figuring correctly (which is highly likely) i figure that we are only using about 15% of the transformers capacity.

If this is true would it be more economical to replace this Tx with a smaller one, or possibly put a small Tx at each machine and one larger one for the single phase? Im curious if we are burning a ton of electricity just by having this big transformer on whether its loaded or not.
My guess is you are not figuring out the loading correctly and how it is distributed from the transformer. Assuming you are talking about all 120 volt loads and @15 amps each, if you had six such loads balanced across all three phases you would only be drawing 30 amps per phase, net on the neutral would be zero if they are all balanced.

You can not just add the amps unless you have a single phase two wire output on the source. you have to consider how the loads are distributed through the transformer. With three phase systems a full balanced load has the square root of three (1.732) factor involved in the calculations, because current is divided three different ways.

If you only have 5- 15 amp 120 volt loads that is 1800VA each or 9000 total VA. - that is only 4% the capacity of a 225kVA transformer.

If your loads were 208 volts instead of 120 then you would have 15600VA - only about 7% of the capacity of a 225kVA transformer.

I was assuming 120 volt loads because you had current on the neutral conductor.
 
I had my maint man work a little on the electrics today. He balanced the single phase loads a little more evenly. No way to get it all perfectly even but it is better and the current on the neutral is half of what it was. I dont understand how that is but then again i dont know much about this stuff. Also our single phase loads were distributed aross both outside legs rather than one single leg as i had thought.
Both outside legs?
The transformer, being Y has three outside legs. Three phases.

One thing that really stuck out to me is the rating on our 208Y transformer. It is rated at 225KVA 3.5% Imp. Its a 480 to 208Y.
That would give you a line current capacity of about 625A. (Assuming 208V line to line.) Possibly greatly more than you need for the CNC machines. But what else does it feed?

Adding up the amp loads from all 3 phases plus the neutral came to about 72 amps. Thats with about 80% of our single phase loads running and no 3 phase machines running. The machines pull around 15amps each and usually have about 5 machines running from this transformer at one time. So that puts us around 150amps theoretically. Unless i am not figuring correctly (which is highly likely) i figure that we are only using about 15% of the transformers capacity.
You simply can't calculate it that way.

If this is true would it be more economical to replace this Tx with a smaller one, or possibly put a small Tx at each machine and one larger one for the single phase? Im curious if we are burning a ton of electricity just by having this big transformer on whether its loaded or not.
Transformer losses are normally quite small. They come in two kinds. Iron loss (Fe) and copper loss (Cu).
The Fe loss is typically less than 1% - for loss calculations I work on 0.7%. The Cu loss might be about 1% at the full rating of the transformer. And it is square law with current. Half the current is quarter of the losses. In your case, if you are running your 225kVA unit at about 10% (1/10) loading, you'd be down to 1% of full load Cu losses. That is 1% of 1%.

In short, your 225kVA transformer losses are pretty unlikely to be burning a ton of electricity. Reducing the rating and/or rebalancing would be, at best, tinkering around at the margins for energy savings.
 
Both outside legs?
The transformer, being Y has three outside legs. Three phases.

I believe he said earlier that this facility once was supplied from an open delta system - so what he meant was the center phase in disconnects and panelboards would have been where the high leg was landed at that time. That was the reason for not placing any single phase loads on the center phase. Nobody changed these single phase loads to balance them when the delta was converted to wye.
 
Status
Not open for further replies.
Top