Separately Derived System from a Separately Derived System

V

Variant

Member
Location
Louisiana
Occupation
Electrician
Greetings. Long time reader, new poster. Sadly, this is on the topic of grounding and bonding of a separately derived system, this time a generator and transformer.. I have a tricky situation with the voltage on a panel fed by a transformer at a commercial location. The panel in question is fed through an ATS to its "main panel." A 3 phase breaker feeds 480 volts to a transformer which feeds 120/208 volts to the trouble panel. The trouble is that the transformer, nor the sub panel are bonded neutral to ground. Of course, this is giving me wild voltage readings of neutral to ground and ground to hot. I very briefly attempted bonding the neutral to ground at the transformer and it did not like it (sparks, immediate signs of over current). The feed "main panel," fed through the ATS is bonded and voltage readings are fine at that point, but everything goes astray at the transformer." There is another, identical transformer in the same room feeding a different panel. but not connected to a generator and the transformer is bonded, thus the voltage readings at the panel are fine. Everything seems to be solidly grounded, but the neutral is floating.
My questions:
In what situation should the transformer, nor the secondary panel not be bonded?
If it is a case of being run through a generator, then why would the primary panel be bonded?
I don't think I am an idiot, but am I being an idiot??
Much appreciated.
 
Variant said:
A 3 phase breaker feeds 480 volts to a transformer which feeds 120/208 volts to the trouble panel. The trouble is that the transformer, nor the sub panel are bonded neutral to ground.
Click to expand...
There should be a system bonding jumper either within the transformer or within the panel depending on where the GEC is connected, is there a GEC? I'm surprised that if the GEC is bonded to the transformer enclosure that the X0 is not actually bonded just not correctly.
 
infinity said:
There should be a system bonding jumper either within the transformer or within the panel depending on where the GEC is connected, is there a GEC? I'm surprised that if the GEC is bonded to the transformer enclosure that the X0 is not actually bonded just not correctly.
Click to expand...
There is a ground wire from the source panel as well as a ground wire to the structure steel, both terminated at the equipment ground in the transformer. As far as the ground in the source panel, it is fed from an exterior disconnect that I have not yet had access to, but voltage and ohm readings at the source panel are normal.
 
Variant said:
There is a ground wire from the source panel as well as a ground wire to the structure steel, both terminated at the equipment ground in the transformer
Click to expand...
The XO from the transformer needs to connect there too.
Variant said:
I very briefly attempted bonding the neutral to ground at the transformer and it did not like it (sparks, immediate signs of over current).
Click to expand...
I don't know what "immediate signs of over current" means. If you tried to bond the XO while everything was live I would expect you to see a spark. I would also encourage you to shut off the 480V breaker feeding the transformer.
 
Variant said:
Greetings. Long time reader, new poster. Sadly, this is on the topic of grounding and bonding of a separately derived system, this time a generator and transformer.. I have a tricky situation with the voltage on a panel fed by a transformer at a commercial location. The panel in question is fed through an ATS to its "main panel." A 3 phase breaker feeds 480 volts to a transformer which feeds 120/208 volts to the trouble panel. The trouble is that the transformer, nor the sub panel are bonded neutral to ground. Of course, this is giving me wild voltage readings of neutral to ground and ground to hot.
Click to expand...

What voltage readings do you get on the 'trouble panel'? Do those voltage readings change with load? Make all the measurements you can (L-L, L-N, L-G and G-N for all phases).

I can think of two possibilities: you have a panel supplied by some sort of autotransformer, not a separately derived system, and you have a panel supplied with a 240V high leg system with some sort of strange tapping.
 
ActionDave said:
The XO from the transformer needs to connect there too.

I don't know what "immediate signs of over current" means. If you tried to bond the XO while everything was live I would expect you to see a spark. I would also encourage you to shut off the 480V breaker feeding the transformer.
Click to expand...
Yes, the spark didn't surprise me, but transformer gave a loud hum and sounds of struggle in accordance with a usual fault/overload of a circuit. Shutting down the power wasn't an option at the time, which is why I made the connection brief. There were also reports of electrical disturbances from parts of the building resulting from this.
 
winnie said:
What voltage readings do you get on the 'trouble panel'? Do those voltage readings change with load? Make all the measurements you can (L-L, L-N, L-G and G-N for all phases).

I can think of two possibilities: you have a panel supplied by some sort of autotransformer, not a separately derived system, and you have a panel supplied with a 240V high leg system with some sort of strange tapping.
Click to expand...
I don't remember the exact voltage readings, but line to line and line to neutral were normal, about 120/208. Neutral to ground was about 180 volts, hot to ground about 320 volts.
 
In my early days of electrical work, the job foreman forgot to bond XO in the transformer for the computers in a Walmart, but he didn’t want to shut off the transformer. He bonded XO live, and as soon as he did, it surged the computer system, and let the magic smoke out of a server! Of course we didn’t let anybody know! LOL! Never bond a transformer live!
 
hillbilly1 said:
In my early days of electrical work, the job foreman forgot to bond XO in the transformer for the computers in a Walmart, but he didn’t want to shut off the transformer. He bonded XO live, and as soon as he did, it surged the computer system, and let the magic smoke out of a server! Of course we didn’t let anybody know! LOL! Never bond a transformer live!
Click to expand...
Thank you for that, I've been wondering if that was too much to do while the equipment was energized. So, did it get bonded while the power was off and was everything ok from there?
 
winnie said:
Those numbers suggest to me that you have a fault from the 480V side to the 208/120V side.
Click to expand...
Now that you say that, I do remember that one of the phases did read 120 volts to ground, while the other 2 read much higher. If there is a fault between sides, wouldn't that make a variable in hot to neutral readings on different phases? Would that cause a problem with bonding?
 
Variant said:
Now that you say that, I do remember that one of the phases did read 120 volts to ground, while the other 2 read much higher. If there is a fault between sides, wouldn't that make a variable in hot to neutral readings on different phases? Would that cause a problem with bonding?
Click to expand...

If you have an ungrounded secondary with a fault from the 480V primary to the secondary, then the H-H and H-N voltages will all be normal but the H-G and N-G voltages will be very wrong.

If the fault is _solidly bolted_ then the voltages can be calculated with some geometric approaches. But if the fault is 'weak' (say a bit of leakage through a thin insulation or through wet dust), then the voltages will be all over the place and very poorly defined.

-Jonathan
 
You must log in or register to reply here.
Top