Transformers in Reverse

[W@ttson]

Hello all,

This might be silly but I am so used to in my head going one way in the usual power distribution that now when I am presented the reverse power flow, its making do a double take.

Typically, we have a service and we envision that to be a Delta to wye. Three phase 4 wire. Any unbalanced current returns on the neutral back to the XO in the transformer. We have a MBJ at the service location and that bond occurs only at that one point (rest are SBJs).

I have a situation where I have a service that is 208V120V 3ph 4 wire. The service/meter / main disconnect switch is about 400ft away as the crow flies from a step up transformer. The transformer is a 3ph customer owned step up transformer, however, it is wired Y to Delta. Y being the 208/120V. There is a N-G in there (I didn't get to see inside the main switch). On the secondary which is the high side is 480V (not 480V/277V, its a delta and no neutral).

Only 3 phases and an EGC is sent from that transformer (no neutral). That feed is ran about 800ft until it hits the building it is servicing. The building has a main disconnect.

That building has a generator and a transfer switch. (The generator has a neutral and it is terminated in the transfer switch but the neutral lug is isolated from everything and since the main service does not have a neutral, there is no intermixing of neutrals there).

I assume that I drive my ground electrodes at the building and just terminate to the main switches ground lug, nothing more?

There are unbalanced Phase to Phase loads on the 480V system. We do see a unbalanced voltages. I assume its from the unbalance? What other concerns are there with running an unbalanced delta as your feed to your loads? The largest loads are 3 ph motors, but they are on seldom. Majority of the time its an odd number of 1 ph heaters.

 

[jim dungar]

Do not connect the source neutral to the X0, wye point, on the 208V side of your transformer.
It will mess with your output regulation.
If the X0 to ground bonding is not removed, you will be regrounding your neutral, which is not allowed by the NEC.

 

[texie]

Looking at your sketch I see all kinds of code and safety issues with this. We can get into the details but for starters you should be using a 208 delta primary X 480/277 secondary transformer for the step up. Are there any line to neutral loads on the 480 system?

 

[W@ttson]

Do not connect the source neutral to the X0, wye point, on the 208V side of your transformer.
It will mess with your output regulation.
If the X0 to ground bonding is not removed, you will be regrounding your neutral, which is not allowed by the NEC.

Can you elaborate on the output regulation issues? There are issues when the large 3 phase loads come on. Some times the voltage sags so much that the generator kicks on.

 

[W@ttson]

Looking at your sketch I see all kinds of code and safety issues with this. We can get into the details but for starters you should be using a 208 delta primary X 480/277 secondary transformer for the step up. Are there any line to neutral loads on the 480 system?

This is existing. Which is why I was scratching my head when I came up to a Y - Delta transformer like that and called it "in reverse". I guess my gut was right, it doesn't seem to pass the smell test.

On the 480V system there are L-N loads but there is a small step down 3PH transformer for those.

What are the immediate safety concerns? The regrounding issue texie mentioned?

I didn't get to look inside the main service disconnect to see if the N-G bond occurred there. I usually never have a Y primary so having the N-G there seemed odd just to start with.

 

[jim dungar]

Oh, don't forget your ground detector scheme on the transformer 480V output.

 

[jim dungar]

Can you elaborate on the output regulation issues? There are issues when the large 3 phase loads come on. Some times the voltage sags so much that the generator kicks on.

Simply put, the connected Wye point will force the primary windings to remain at 120V which prevents the 480V from truly accommodating an unbalance loading, like you may experience when starting large motors.

I do not know of a single transformer manufacturer that does not warn against connecting the Wye when a transformer is run in reverse.

 

[W@ttson]

Simply put, the connected Wye point will force the primary windings to remain at 120V which prevents the 480V from truly accommodating an unbalance loading, like you may experience when starting large motors.

I do not know of a single transformer manufacturer that does not warn against connecting the Wye when a transformer is run in reverse.

Very interesting! I read about backwards wiring transformers a long time ago. I never ran into it again until today. I will revisit that literature.

Thank you

 

[W@ttson]

Ok so here it gets a little weirder.

The name plate actually calls out the primary as 208/120V

 

[texie]

Ok so here it gets a little weirder.

The name plate actually calls out the primary as 208/120V

View attachment 2568850

This looks like some type of utility transformer as it is oil filled that has been repurposed for your job. Not really my area of expertise as to what this was bait for. But for your application it is totally the wrong thing to use for many reasons.

 

[W@ttson]

This looks like some type of utility transformer as it is oil filled that has been repurposed for your job. Not really my area of expertise as to what this was bait for. But for your application it is totally the wrong thing to use for many reasons.

yes, it is exactly an oil field utility looking transformer that is customer owned.

 

[W@ttson]

So I took some voltage out there on the 208V side and on the 480V side:

208V side:

AB|AC|BC
202V|204V|206V

AN|BN|CN
117V|118V|119V


==========================================

480V side:
AB|AC|BC
469V|465V|474V

AG|BG|CG
271V|275V|269V


What I am seeing is that this is an ungrounded system since it is not a corner grounded Delta, or a high leg delta.
The anomolies with the voltages are likely due to that XO bond of the neutral and the gnd.
It is interesting that even though the "ground" that is sent out and is connected to the XO along with that neutral, is still measuring ~277V.

 

[ptonsparky]

So I took some voltage out there on the 208V side and on the 480V side:

208V side:

AB|AC|BC
202V|204V|206V

AN|BN|CN
117V|118V|119V


==========================================

480V side:
AB|AC|BC
469V|465V|474V

AG|BG|CG
271V|275V|269V


What I am seeing is that this is an ungrounded system since it is not a corner grounded Delta, or a high leg delta.
The anomolies with the voltages are likely due to that XO bond of the neutral and the gnd.
It is interesting that even though the "ground" that is sent out and is connected to the XO along with that neutral, is still measuring ~277V.

I think that 277 will change as 480v loads change. Say 480v lighting circuits. IDK enough to explain it. Capacitance?
(Some one usually tells me when I have my head where it's dark.)

 

[jim dungar]

The secondary L-G voltage is due to 'coupling capacitance' and it is primarily caused by the secondary conductors.

Ground is ground, there is no such thing as primary or secondary ground. There are however primary and secondary bonding jumpers, and there should not be one on your primary.

Is this a new use of an old transformer and as such subject to the NEC? Several code cycles back, language was added to the NEC saying you could not reverse feed a transformer that had a nameplate that said Primary and Secondary.

 

[W@ttson]

The secondary L-G voltage is due to 'coupling capacitance' and it is primarily caused by the secondary conductors.

Ground is ground, there is no such thing as primary or secondary ground. There are however primary and secondary bonding jumpers, and there should not be one on your primary.

Is this a new use of an old transformer and as such subject to the NEC? Several code cycles back, language was added to the NEC saying you could not reverse feed a transformer that had a nameplate that said Primary and Secondary.

This is an old transformer. My task was to add the ground electrodes at the remote separate building/structure. I then started to investigate upstream because I was expecting there to be a neutral at the building and there wasn’t. I landed on this transformer.

It is an existing transformer that the client expects to use as long as possible.

Is that is it really back feeding if the “primary” is the 208v/120v side?

I have no choice but recommend that transformer be replaced. Have a typical delta wye installed. Dead end the neutral if they want, but we need more consistent voltage operation and as of right now I don’t see a proper way to safely clear a L-G fault.

 

[jim dungar]

An old piece of equipment is still subject to the current NEC when it is being reused in a new installation.

Yes, it is reverse feeding when the input/primary to the transformer is different than what the nameplate says. Many modern transformers now only say HV and LV for this reason.

Adding a "grounding transformer" is also an option for providing L-G fault protection on the 480V side, but a new 75kVA step-up transformer is probably a similar cost.

 

[electrofelon]

I would say you have two issues:

1. The wye side XO being bonded and a neutral run to it. Very bad things can happen if you have a major imbalance, say particularly of you lose a phase. I had this exact thing happen and a hole blasted through the 2-in EMT that supplied the transformer when it faulted. Of course this is an easy fix just remove the neutral and remove the bonding strap if present.

2. The 480 side being ungrounded. Not really a safety concern in theory, but it's more about unqualified people or people who just will assume it's grounded and don't understand ungrounded systems. Also it appears you do not have any ground detection system.

You might be able to help sell a new proper transformer by discussing the transformer losses. If you bump up to a new 2016DOE transformer. I bet your losses will be quite a bit less being that unit was manufactured in 89, and it will eventually pay for itself.

 

[W@ttson]

I would say you have two issues:

1. The wye side XO being bonded and a neutral run to it. Very bad things can happen if you have a major imbalance, say particularly of you lose a phase. I had this exact thing happen and a hole blasted through the 2-in EMT that supplied the transformer when it faulted. Of course this is an easy fix just remove the neutral and remove the bonding strap if present.

2. The 480 side being ungrounded. Not really a safety concern in theory, but it's more about unqualified people or people who just will assume it's grounded and don't understand ungrounded systems. Also it appears you do not have any ground detection system.

You might be able to help sell a new proper transformer by discussing the transformer losses. If you bump up to a new 2016DOE transformer. I bet your losses will be quite a bit less being that unit was manufactured in 89, and it will eventually pay for itself.

The neutral is definitely landed on the “primary” and the strap is definitely there.
See picture below.

 

[electrofelon]

The neutral is definitely landed on the “primary” and the strap is definitely there.
See picture below.

Remove it! Run!

 

[wwhitney]

Yes, it is reverse feeding when the input/primary to the transformer is different than what the nameplate says.

But the nameplate shown does have the primary/secondary labeling matching the current usage.

Of course, if the transformer appears to have originally been a step-down transformer, that suggests the current nameplate is not the OEM nameplate, which raises other questions.

Cheers, Wayne