True power isolation

Status
Not open for further replies.
S

suemarkp

Senior Member
Location
Kent, WA
Occupation
Retired Engineer
This is an excerpt from a government document explaining the requirements for facility power for isolation transformers. It is old and note the reference to NEC 250-5 which doesn't seem to have a direct map to the more recent NEC numbering style. I think it is referring to some section on separately derived systems.

Here's the quote: "It should be installed per NEC 250-5(d) as a separately derived system and, as such, establishes a new fault protection subsystem. It should not be tied to the green wire serving the power main side, as this defeats the intent of both the NEC for protection and the use of this type of transformer to breaker ground loops. Further, some isolation transformers can be designed with the ground and neutral conductors being common to the primary and secondary windings. This reduces the siolation effectiveness of the transformer."

So I understand wanting a transformer secondary whose neutral is not common to the input neutral (if it even has an input neutral). However, is there any legal way to not ground the secondary, or ground it in such a way that it doesn't work back to the primary side "green wire"? I don't see how when this is installed in an industrial building with building steel being used as the grounding path (this is the electrode present throughout the building and I would expect it to be electrically continuous).
 
suemarkp said:
However, is there any legal way to not ground the secondary, or ground it in such a way that it doesn't work back to the primary side "green wire"? I don't see how when this is installed in an industrial building with building steel being used as the grounding path (this is the electrode present throughout the building and I would expect it to be electrically continuous).
Click to expand...

I believe you would have 250.30(B). The neutral wouldn't be tied to a GEC.
 
Am I permitted to have an ungrounded system (250.30(B)) from a SDS in premises power system (this will be a 208/120V 150 KVA system)? Wouldn't that violate 250.21(3) since I don't meet any of the criteria in (a) - (d). But all of this seems to conflict with 250.20(D) which sends you to 250.30 making me think it is allowed??
 
Last edited:
suemarkp said:
Am I permitted to have an ungrounded system (250.30(B)) from a SDS in premises power system (this will be a 208/120V 150 KVA system)? Wouldn't that violate 250.21(3) since I don't meet any of the criteria in (a) - (d). But all of this seems to conflict with 250.20(D) which sends you to 250.30 making me think it is allowed??
Click to expand...

I have never personally done this, but from what I read it seems that it is allowed. I could be missing something here. I'm not sure how but you may be able to justify 250.21(5). But in any case, the biggest thing with isolation transformers are the secondary windings of an iso-transformer have no physical electrical connection to the primary windings.
 
suemarkp said:
This is an excerpt from a government document explaining the requirements for facility power for isolation transformers. It is old and note the reference to NEC 250-5 which doesn't seem to have a direct map to the more recent NEC numbering style. I think it is referring to some section on separately derived systems.

Here's the quote: "It should be installed per NEC 250-5(d) as a separately derived system and, as such, establishes a new fault protection subsystem. It should not be tied to the green wire serving the power main side, as this defeats the intent of both the NEC for protection and the use of this type of transformer to breaker ground loops. Further, some isolation transformers can be designed with the ground and neutral conductors being common to the primary and secondary windings. This reduces the siolation effectiveness of the transformer."

So I understand wanting a transformer secondary whose neutral is not common to the input neutral (if it even has an input neutral). However, is there any legal way to not ground the secondary, or ground it in such a way that it doesn't work back to the primary side "green wire"? I don't see how when this is installed in an industrial building with building steel being used as the grounding path (this is the electrode present throughout the building and I would expect it to be electrically continuous).
Click to expand...
There is no code compliant way to avoid a connection to the primary grounding conductor if the output of the secondary is a voltage that is required by 250.20 to be a grounded system.
 
don_resqcapt19 said:
There is no code compliant way to avoid a connection to the primary grounding conductor if the output of the secondary is a voltage that is required by 250.20 to be a grounded system.
Click to expand...

Yet it appears that there once was.

The problem with the SDS required to be conected to the primary ground is that it means you cant have a truly isolated SDS, which under certain conditions, is what you need for safety.

Somewhere along the line, with the best of intentions, the code making panel got it wrong.

There is one exception, which is for OR sockets with ground leak detection...
 
Thanks for the comments. The government document is for TEMPEST/EMSEC requirements. In these installations, there is usually an earthed equipotential grid which is used to interconnect chassis of equipment. But I think too many pieces of equipment connnect their EGC to the chassis, so the whole concept isn't going to work very well since the EGC makes it way back to the building common ground electrode system. It may be better, since the lowest impedance paths are near the SDS, but the whole building is still going to be interconnected.

Just wanted to make sure I understand things and don't spend too much on a special transformer which is going to be defeated by required grounding anyway.
 
From the handbook, 240 volt and 480 volt 3 phase, 3 wire systems are examples of systems that don't need to be grounded.

There are also some other exceptions, like some control power circuits, some furnaces, some circuits in health care facilities....

Steve
 
OK this is a subject my career revolves around. When we install a SDS, and lets stick to a transformer if you are going to debate me OK? When you install a SDS you create a new ground reference point, and thus break all so called ground loops, noise, etc.... It makes no difference from a safety or operational POV if the if the Xo is part of the faiclity GES or isolated from it because of the requirement of the bond of the EGC primary to the case of the transformer.

Let the debate begin.:grin:
 
Last edited:
By building a SDS you do indeed create a new ground reference point, but the old reference point still exists and is still connected tothe load side of the SDS because of the bond to the primary side ground, and you can thus have currents flowing in the ground from the primary side. So you dont get rid of the so-called ground loops.

This is all down to the multi grounded neutral thing, where there are potential gradients all over the place. An isolating transformer offers the opportunity to "start again", which for some applications is exactly whats needed (my favorite - cow milking sheds) but the reqirement to bond the new ground to the old ground defeats that purpose.
 
Status
Not open for further replies.
Top