• We will be performing upgrades on the forums and server over the weekend. The forums may be unavailable multiple times for up to an hour each. Thank you for your patience and understanding as we work to make the forums even better.

Grounding SDS back to substation

Merry Christmas

holmessph

Member
Location
Ohio
Occupation
Data Center Tech
Transformer 1 and 2 are both separately derived and each would have its own N to G bond plus a GEC run from each main or system bonding jumper, common busbar, to the facility (common contiguous) grounding electrode system.

Then you describe loading 1's loads onto 2, which could be described typically as a dual source bus with a tie, and you say the majority data center loads and configured line to neutral. You would have high neutral current with two (or more) N to G bonds putting your high neutral current on the EGC and GEC paths.

If the source transformers are remotely located from the switchgear bond, it is possible or likely that also has another N to G bond, with two such source making four minimum have to be checked. There should be only one if at all possible.

This is one of the reasons, this type of data center dual source bus, would only have line to line connected loads at the dual bus, where the system bonding jumper is, because of the possibly unavoidable multiple N to G bonds. No neutral connected loads is one way to clean it up, but not possible in your case.

Anyway, you have identified where to look.

I would recommend starting at square 1. Look at what you have on paper and decide where your one main or system bonding jumper with its associated GEC conductor should be. Then examine each transformer and downstream switchgear for their present or not N to G points. Survey knowing what you are looking for before proceeding.

Did not read the entire thread.

This morning I was able to verify my assumption.

So even though we’re not using switch 1 to service any load, it’s still hooked up, and the G+N are bonded. The same bus bars then go over to switch 2 from the transformer and in that switch also has G+N bonded.

What I believe is happening is we have a parallel path in the same SDS.

The following drawing is now verified as accurate and this more than likely explains what we’re seeing.

IMG_6803.jpeg
 

holmessph

Member
Location
Ohio
Occupation
Data Center Tech
The only thing that is outside is the main substation which is 69kV to 4160v.

The disconnects and everything else are all indoors.

The first disconnect listed is 300+ FT away from the substation, the second disconnect is another 140FT down the line.

The two transformers and the switchgear are within 15FT of each other.
 
The only thing that is outside is the main substation which is 69kV to 4160v.

The disconnects and everything else are all indoors.

The first disconnect listed is 300+ FT away from the substation, the second disconnect is another 140FT down the line.

The two transformers and the switchgear are within 15FT of each other.
Thanks. Still just trying to fully digest everything. The 4160, is it grounded or ungrounded? If it's grounded, is the grounded conductor run with the phases? Is there a bonding conductor run?
 

holmessph

Member
Location
Ohio
Occupation
Data Center Tech
Thanks. Still just trying to fully digest everything. The 4160, is it grounded or ungrounded? If it's grounded, is the grounded conductor run with the phases? Is there a bonding conductor run?

If you take a look at the 1-line I posted above it might answer some of those questions.

1) The 4160v primary side has a ground conductor that goes all the way back to the substation thru 2 non-fused disconnects. The substation outside has a grounding grid which I presume is where the entire outside substation is grounded too.

2) The secondary side of the transformer has a ground connection via a bus to the ground conductor in the MV Disconnect right next to it, and then goes into the switchgear attached to it, where the N+G bond exists.. Important to note, that switch gear is not in use, it's hooked up but it's not in use (although it was intended to be used in the original design). Neither switch, and neither transformer have a ground to a building column or directly to a local rod, so we're already in violation there, but as others pointed out that's likely not causing why we soo so much current going over the ground that shouldn't be.

3) The active 4160v to 480v transformer has a bus for each phase, N and G, in addition to being fed into switch 1, there's additional conductors running to switch 2. Switch 2 ALSO has a G+N bond inside of it. The load we're serving IS connected to switch 2. I know this is an odd situation, but the transformers and switch gear are used. When we wired everything in, we found that transformer 2 had some issues during turn up that would need addressed before loading up, so we back fed switch 2 with feeds from transformer 1.

My assumption here based off of what I have been able to verify is that by not unhooking the switchgear connected to transformer one, we have a system that's bonded twice. Once in switch 1 and again in switch 2, and from switch 2 we then service a load. So the question, is, would this not cause a parallel ground path and thus a loop? This would seem to be the case, no different than if you bond N+G on a sub-panel in a house.
 
Ok thanks. However you didn't answer one of my questions. Is the 4160 a grounded or ungrounded system? Is the transformer on the 4160 side a delta? (Although that itself does not determine whether the 4160 is grounded or not). I'm not sure yet if the 4160 system being grounded or not matters, it's just nice not to have variables when you're mulling over something.
 

holmessph

Member
Location
Ohio
Occupation
Data Center Tech
Ok thanks. However you didn't answer one of my questions. Is the 4160 a grounded or ungrounded system? Is the transformer on the 4160 side a delta? (Although that itself does not determine whether the 4160 is grounded or not). I'm not sure yet if the 4160 system being grounded or not matters, it's just nice not to have variables when you're mulling over something.

It's a 4160v to 480v (Delta to WYE) grounded system. The entire building, even the substation outside is a grounded system, there is no ungrounded delta running around here anywhere, which I've seen before in these older buildings.
 
It's a 4160v to 480v (Delta to WYE) grounded system. The entire building, even the substation outside is a grounded system, there is no ungrounded delta running around here anywhere, which I've seen before in these older buildings.
Ok so the 69-4.16kv transformer has a wye secondary with the XO bonded? So you have a 4160y/2400 uni-grounded system (MV neutral is bonded only at the substation)? Three phases plus a ground / Bond wire run from the substation to you?
 
OK, there are two N-G bonds close together and no connection to a GES anywhere? (Or is there a GES at the transformer?)

We still get back to the issue that the SDS needs to connect to a GES, and the EGC conductor in the structure needs to connect to a GES.
 
I reread everything and have an additional thought. The problem with lifting that bond in the unused switchgear is you have live conductors to that gear and if you lift the N-G bond you wont have a good fault path back to the source. I am guessing any fault current will get over to the other switchgear through conduits and what not, and then to the neutral and back to the source by the N-G bond there, but it really shouldnt be that way. Seems like you should have the N-G bond at the transformer, and isolated neutral at the switchgears.
 

__dan

Senior Member
Both transformers 1 and 2 would be expected to have the main or system bonding jumper (required installed) somewhere between the physical transformer and the first disconnect. If you had two load side mains and a tie in the same cabinet assembly, one jumper on the common grounding busbar may be possible and that point is also where the EGC and GEC would connect to.

But you do not describe two mains with a tie in the same switchgear.

You describe transformers 1 and 2 remote from each other, but each close to their own load side main, where the system bonding jumpers and common connection point to the facility EGC and GEC would be.

That would make two such N to G connections necessary. I don't know if there is an interpretation where you could use 2's system bonding jumper and common grounding busbar to serve 1, as that would not be between 1's transformer and its load side, first disconnect. It may be possible idk but I am not looking at the code for it.

If 1, the transformer with the problem, is *never* connected to a dual source bus with 2, it would not be an issue. You can disconnect it.

If 1 were to be put online, then it would need its (possibly own) system bonding jumper and GEC, EGC, then solidly connecting to 2's bus, you would be back to where you are right now, with two N to G connections.

I have seen it worse and no one knew what was going on. But the original design engineer was *very* good and loaded all the data center loads line to line so there was (no) neutral current (from that). However still, no one had grasped this insight or knew what they were doing and the entire facility lighting was line to neutral 277 straight off the dual source bus (triple source actually with diesels gensets there also, (5) 1 megaWatters each). With additional N to G bonds at the outside switchgear.

It seems to me the choice to arrange the data center loads as line to neutral was ill advised. Why would they do this, saving one breaker pole and panelboard slots ???

Typical datacenter has (large) UPS and generators also, then a large footprint where they are physically remote from each other. Many opportunity for both missing N to G bonds and at the same time, extra N to G bonds.

The only time I have seen this done cleanly was the long gone brilliant design engineer saw this coming and loaded everything at the dual source bus line to line, zero line to neutral loads. Then after he was gone they loaded stuff line to neutral or up until the breakers tripped.

I think you would want to work it out on paper first. Determine where you want to be, then look at where you are, before proceeding.
 

__dan

Senior Member
If you would ground transformer 1 using transformer 2's system bonding jumper, you would effectively be declaring transformer 1 is "not separately derived".

This is done all the time, properly, with gensets.

However, doing this with two otherwise identical dual transformers would, imo, be a unicorn.

Even if you ran into another person to even entertain thinking about it, being nonstandard is the problem. Guys only know and want to deal with what they are familiar with. It is very predictable that there will be some future arguement, transformer 1 "is" or "is not" separately derived, and "does", "does not", need its own system bonding jumper.

You can see you are there yourself, having stated the problem was the missing bond to the building steel.

Declaring 1 "not separately derived", is easy to show you are code complaint. Getting someone else to understand what you want to do and agree with you, for reason of code compliance and objectionable current, that would be very predictably "difficult".

The predictable path, from experience, highly probably the guys would just live with the objectionable current and keep resetting the breaker. It is that bad out there.
 

__dan

Senior Member
Moving the SBJ's to the transformer is quite easy yes?
Yes, but with a 4 wire system you (would not want to) switch the neutrals also.

That means 3 pole switches and solidly connected neutrals, in a scenario with high neutral current and neutral loading. Where does the "one" SBJ go.
 

holmessph

Member
Location
Ohio
Occupation
Data Center Tech
I reread everything and have an additional thought. The problem with lifting that bond in the unused switchgear is you have live conductors to that gear and if you lift the N-G bond you wont have a good fault path back to the source. I am guessing any fault current will get over to the other switchgear through conduits and what not, and then to the neutral and back to the source by the N-G bond there, but it really shouldnt be that way. Seems like you should have the N-G bond at the transformer, and isolated neutral at the switchgears.

The electrical engineer who designed this setup finally called me back. He says that having each switchgear have it's own bond is fine since the transformer is not bonded N+G. He said then each switchgear is it's own separately derived system. So I am getting conflicting data in that regard.
 

holmessph

Member
Location
Ohio
Occupation
Data Center Tech
Yes, but with a 4 wire system you (would not want to) switch the neutrals also.

That means 3 pole switches and solidly connected neutrals, in a scenario with high neutral current and neutral loading. Where does the "one" SBJ go.

According to the electrical engineer on this job, he's stating that NOT using a bonding jumper at the transformer, and instead doing it at each switchgear connected to the transformer secondary makes each switchgear it's own separately derived system. He claims that's the way it should be...

He is also claiming that it's not abnormal to see current going over the ground. Either (a) in a non-linear scenario like ours where we're using a lot of phase to neutral current and (b) even in a 3-phase 480v roof fan he claims we will see leakage to the ground. I have conflicting opinions on this from multiple electricians lol
 

wwhitney

Senior Member
Location
Berkeley, CA
Occupation
Retired
According to the electrical engineer on this job, he's stating that NOT using a bonding jumper at the transformer, and instead doing it at each switchgear connected to the transformer secondary makes each switchgear it's own separately derived system. He claims that's the way it should be...
He is WRONG. Having two switchgears connected to ONE transformer is NOT two separately derived systems. It is one SDS with two sets of transformer secondary conductors. Maybe he is not aware the way it is currently configured?
 
Top