3 phase transformers

[Jody Boehs]

Hello,

I have installed a 45 kVA 480V 3ph / 240V 3ph w/120V CT. This particular transformer was installed, not for the need of 240V 3ph, but for 120V 1ph. The reason I went with a 3ph transformer was to keep an equal load on all 3 legs of our 3ph 4 wire system (I was told to do this by my boss at that time). Needless to say I now have a new boss that trusts my knowledge in installing what is appropriate. However, I have come to a point where I am needing to install another transformer that is identical to the first one I explained (I already have it on location).

So, here's my questions...

In reading the specs for this 45 kVA transformer, it gives these... "Installation Notes. CAUTION: When using the 120V center tap for single-phase applacations, the single-phase load should not exceed 5% of the three-phase kVA rating. Connect the X3 "high leg" tot he "B" phase per NEC 384-3 (do not use X3 leg for 120V lighting)."

This has caused 3 questions for me.

1. I did not know this, and I've already installed 1 of these solely for the 240/120 1ph. I connected all 3 legs on the primary side and only X1, X2, and X4(center tap). X3 was left unterminated. This particular transformer is supplying 4 - 20A 120V circuits, each of which rarely(if at all) will pull 20A and a drill press which operates on 240V at approx. 9.5A. So, am I OK in this setup? Hopefully so, because to explain that I need to change all that equipment out will go over like a lead balloon.

2. I have an indentical transformer as the one described above (just got it delivered yesterday). My plan was to wire it the same way. It will be supplying 6 - 120V 20A circuits, again each of which will never pull the full 20A. I'd be surprised if this system pulls 20A total across all circuits combined. Am I OK to continue with this plan? OR should I redesign and start over? I think I'll be OK, but would like to know what you more experienced folks think.

3. I will be installing another system like the ones described above, only a tad bit bigger. The plan was to use a 75 kVA transformer (same configuration as previous transformers). However, this system was to supply 3 welders and a plasma cutter (4 - 50A circuits), and several 120V circuits for grinders and sanders. These circuits obviously will pull a whole lot more amperage. I definitely will be exceeding the 5%. What should I do here? Redesign and get a different transformer?

I need some help here...

Thanks in advance!

 

[jim dungar]

If you do not use 3-phase, then you do not need 3-phase transformers.

As you describe your installation, you are energizing the core and coils for all three windings, but you are only using one of them.
Using only a single secondary winding of a delta connected transformer does not provide a balanced load across all of the primary phases.

The warning about 5% loading on the CT winding is assuming that you are also loading the transformer on the other two windings (i.e. something is connected to the X3 terminal).
You need to check with your transformer manufacturer to see how much 1-phase only loading you can have, but I would be surprised if it was anything other than 1/3 (i.e. 15kVA [MENTION=128116]120/240[/MENTION]V) of the full load rating.

 

[kwired]

I agree with Jim in that if a goal was balancing the main lines, you won't accomplish that goal if all/most the load is 120 volt loads, and that a wye secondary would probably better suit the needs.

 

[iceworm]

....
1. I did not know this, and I've already installed 1 of these solely for the 240/120 1ph. I connected all 3 legs on the primary side and only X1, X2, and X4(center tap). X3 was left unterminated. This particular transformer is supplying 4 - 20A 120V circuits, each of which rarely(if at all) will pull 20A and a drill press which operates on 240V at approx. 9.5A. So, am I OK in this setup? Hopefully so, because to explain that I need to change all that equipment out will go over like a lead balloon. ....

5% of 45kva = 2.25kva which gives 2250/240 = 9.4A Neglecting the receptacles as little used, I guess this one is okay.
As for explaining the need for changing design: Yes, that is what happens. I don't like explaining those either.

.... 2. I have an indentical transformer as the one described above (just got it delivered yesterday). My plan was to wire it the same way. It will be supplying 6 - 120V 20A circuits, again each of which will never pull the full 20A. I'd be surprised if this system pulls 20A total across all circuits combined. Am I OK to continue with this plan? OR should I redesign and start over? ....

Probably okay. However, I always feel like I am leaving behind a trap for the next person. I have thought about a placard "Do not exceed 2200 KVA continuous" But that sucks. I can hear the boss, "We put this in for the Admin's computer and printer. We both know she will plug in a heater. You're telling me that will overload the xfm?"

As for calling the xfm mfg, you already have the answer:

"Installation Notes. CAUTION: When using the 120V center tap for single-phase applacations, the single-phase load should not exceed 5% of the three-phase kVA rating. Connect the X3 "high leg" to the "B" phase per NEC 384-3 (do not use X3 leg for 120V lighting)."​

What is the mfg going to say? "Oh yeah, we lie in the Installation Notes. They can be loaded way higher than that." Probably not.

(edit to add) Reread Jim's comment about the mfg assuming load on the other phases. So maybe I'm in error. I am guessing the mfg is alluding to imbalance. But I don't know that.

...
3. I will be installing another system like the ones described above, only a tad bit bigger. The plan was to use a 75 kVA transformer (same configuration as previous transformers). However, this system was to supply 3 welders and a plasma cutter (4 - 50A circuits), and several 120V circuits for grinders and sanders. These circuits obviously will pull a whole lot more amperage. I definitely will be exceeding the 5%. What should I do here? Redesign and get a different transformer? ...

Redesign and get a different transformer?
Resounding definite - Maybe
Assumption: The 50A circuits are 1phase, 240V
Map out the loads (Load study). Spread the 50A circuits around the three phases. Pick the lightest one for #4 and put it on the CT phase. See how you match upon the imbalance. If you are over the 5%, make a judgement call on the duty cycle.

If you don't like it, consider getting a 208/120 xfm. Check the equipment nameplates to se if 208V is okay.

If the equipment does not like 208V, consider two 1ph 120/24 transformers.

Welcome to being the engineer of record. We all know design engineers are pinheads. Hopefully the hat won't pinch too much.


the worm

 

[kwired]

As for calling the xfm mfg, you already have the answer:

"Installation Notes. CAUTION: When using the 120V center tap for single-phase applacations, the single-phase load should not exceed 5% of the three-phase kVA rating. Connect the X3 "high leg" to the "B" phase per NEC 384-3 (do not use X3 leg for 120V lighting)."​

What is the mfg going to say? "Oh yeah, we lie in the Installation Notes. They can be loaded way higher than that." Probably not.

(edit to add) Reread Jim's comment about the mfg assuming load on the other phases. So maybe I'm in error. I am guessing the mfg is alluding to imbalance. But I don't know that.

Be nice if they would be a little more specific on what that 5% applies to.

The 120/240 side of the delta is basically the same thing in OP's case as a 15 kVA single phase transformer secondary.

should be able to load each half of that side to 7.5 KVA without overloading anything. If both sides are loaded to same level neutral conductor is carrying zero current.

If they are talking about supplying 208 to neutral loads - not impossible but not something you normally see either, I can see that being tricky on determining what current goes where in a full delta, open delta it has no choice of where to go but through one and a half windings.

 

[jim dungar]

Reread Jim's comment about the mfg assuming load on the other phases. So maybe I'm in error. I am guessing the mfg is alluding to imbalance. But I don't know that.

The historical 5% warning is based on the transformer supplying a mix of 3-phase and 1-phase loading. Circulating currents can also become an issue with unbalanced loads on delta transformers, especially with 120V loads on just one half of a winding.
It is pretty dependent on how the transformer core and coil assembly is constructed, for example a five legged core would behave differently than a legged core and a triplex core (individual cores in a single housing) would be different still.

Some manufacturers no longer include this restriction and expect that you will reduce the 3-phase loading as you increase the 1-phase load. However, I feel it is still best to simple use an open delta configuration to supply 240/120V 3-phase 4-wire loads, and single phase transformers for just single phase loads.

 

[iceworm]

The historical 5% warning is based on the transformer supplying a mix of 3-phase and 1-phase loading. Circulating currents can also become an issue with unbalanced loads on delta transformers, especially with 120V loads on just one half of a winding.
It is pretty dependent on how the transformer core and coil assembly is constructed, for example a five legged core would behave differently than a legged core and a triplex core (individual cores in a single housing) would be different still.

Some manufacturers no longer include this restriction and expect that you will reduce the 3-phase loading as you increase the 1-phase load. However, I feel it is still best to simple use an open delta configuration to supply 240/120V 3-phase 4-wire loads, and single phase transformers for just single phase loads.

Jim -
As for the historical 5% warning, I would really hope the xfm mfg would not include it in the installation instructions, or specifications unless it was true and needed to be controlled. If the mfg are just throwing the warning in cause they always have - that's pretty dim.

Q1: Assuming 3ph - 4W means three phases and a neutral -
I have never seen a load that required 240V, 3phase Delta and a neutral. What would be an example?

I have seen equipment that required a 240V grounded supply (these were vfds). Those were met with 240/139 grounded Y - never 240D, grounded center tap

Q2: Why would you use Open-Delta as opposed to Delta?

Personal opinion: I would spec 208Y, unless the alien containment field required 240D. That one would get a dedicated transformer. And, as you said, put in 1ph xfm for the 1ph loads.

 

[Jody Boehs]

Redesign and get a different transformer?
Resounding definite - Maybe
Assumption: The 50A circuits are 1phase, 240V
Map out the loads (Load study). Spread the 50A circuits around the three phases. Pick the lightest one for #4 and put it on the CT phase. See how you match upon the imbalance. If you are over the 5%, make a judgement call on the duty cycle.

If you don't like it, consider getting a 208/120 xfm. Check the equipment nameplates to se if 208V is okay.

If the equipment does not like 208V, consider two 1ph 120/24 transformers.

Welcome to being the engineer of record. We all know design engineers are pinheads. Hopefully the hat won't pinch too much.


the worm

Yeah...I'm not that smart and, by George, I AM NOT AN ENGINEER!! HAHAHA

But anyhow... SO your recommendation would be to go with a 1ph xfm? I thought a transformer was a constant load?....Hence the reason I and my previous boss thought we should balance the load across all 3 legs.

If the equipment accepts 208V then i should get a 480 primary to 208/120 secondary 3ph xfm?

If not, then just use a 480 to 240/120 1ph xfm?

Am I reading you right?

 

[ActionDave]

Y

If the equipment accepts 208V then i should get a 480 primary to 208/120 secondary 3ph xfm?

If not, then just use a 480 to 240/120 1ph xfm?

Am I reading you right?

That is correct

 

[iceworm]

(edit to add) Slow poster - following Dave's comments:

Yeah...I'm not that smart and, by George, I AM NOT AN ENGINEER!! HAHAHA
You are today. This is system design.

But anyhow... SO your recommendation would be to go with a 1ph xfm? I thought a transformer was a constant load?....Hence the reason I and my previous boss thought we should balance the load across all 3 legs.
You are going to have three transformers. 2 - 45kva and 1 - 75 kva. If there is a problem with the 480 being too far unbalanced, you may have to go back and roll phases on the installed small xfms to balance it on the 480V side.

If the equipment accepts 208V then i should get a 480 primary to 208/120 secondary 3ph xfm?
That would be my first recommendation. Probably Jim's as well - but I won't skeak for him

If not, then just use a 480 to 240/120 1ph xfm?
Yep - that sucks but about the best you have for what you are working with.

Am I reading you right?

Close enough

the worm

 

[kwired]

Q1: Assuming 3ph - 4W means three phases and a neutral -
I have never seen a load that required 240V, 3phase Delta and a neutral. What would be an example?

I don't think you will find any individual loads that need all three phases and a neutral. What you can run into is three phase loads as well as line to neutral loads supplied by the same source, whether it be service or separately derived source.

 

[jim dungar]

As for the historical 5% warning, I would really hope the xfm mfg would not include it in the installation instructions, or specifications unless it was true and needed to be controlled. If the mfg are just throwing the warning in cause they always have - that's pretty dim.

The 5% level is valid. It has to do with circulating currents in the delta, the amount of three phase loads and a severely imbalanced (not equal on both sides of neutral) single phase loading. Effectively you can overload a transformer simply by turning off all the loads fed from 1/2 of the center tapped winding.

Effectively, by ignoring power factor, the formula for the center tapped winding is 2.5S + T. So a 75kVA 3-phase unit would be fully loaded with 30KVA of balanced single phase load and no three phase load at all. But this ignores the circulating current inside the transformer caused by the voltage regulation for each of the three coils. A 1.24% voltage difference would result in roughly an 8% circulating current, so the transformer would now be overloaded. We have to consider the increased losses of the transformer caused by the overloading.

It all depends on how the transformer core is built. This is why I suggested contacting the transformer manufacturer with the specific load profile.
Wye-delta and open-delta configurations do not have the circulating current problem.

 

[iceworm]

.... open-delta configurations do not have the circulating current problem.

Ah - yes. And that would be an excellent reason to use an open delta as opposed to closed delta.