Transformers Delta / Wye

[iwire]

Let's say you go to a common GEC busbar and amp clamp the conductors terminating there. You see 5, 7, 9, 12, 23, ... amps each conductor terminating there. Is it objectionable? Is it a violation? This would assume the wiring itself is compliant.
.

Nowhere enough info to say.

Could be fine, could be a problem, might even be hazardous.

But just because there is current on an EGC does not make it automatically objectionable current.

In some instances size of the signal may not matter, if it interferes with normal operation.

Again, the equipment is designed poorly if the 'noise' on EGC supplying it causes an issue.

Even if it does occur in some manner, there is a way to construct the GEC system in a way that would meet someone's opinion of best practice

This is they type of thinking that leads to voodoo grounding schemes such as triad ground rods etc.

 

[kwired]

Are all these transformers installed in same vicinity or at least along different points along a common path for the feeds to them, or are they more scattered around the facility each on it's own path from the source?

The common 3/0 works great when you have one common path to most of them, otherwise just run necessary individual GEC's back to the service or main building supply and bond to the GES there.

 

[Buck Parrish]

Okay, thanks. ..
They're scattered through out all on one floor on a slab. The farthest is about 275 ft away.

 

[__dan]

Are all these transformers installed in same vicinity or at least along different points along a common path for the feeds to them, or are they more scattered around the facility each on it's own path from the source?

The common 3/0 works great when you have one common path to most of them, otherwise just run necessary individual GEC's back to the service or main building supply and bond to the GES there.

I am thinking there may be a way to put in ground rods at each location and then the run to the facility GES system is an electrode bonding jumper and not a GEC. The rods by themselves would do nothing but the savings may be in a smaller and less restricted run of wire. The electrode bonding jumper may be an easier less costly install compared to the equivalent GEC.

I have not checked this with the codes references I would need if I wanted to justify this. It is something I would consider looking at.

 

[__dan]

You are talking about specific, unusual installations to buttress your point.

None of that is typical.

I am not going to disagree with you there. What I have seen is gifted contractors happily doing an install several times to fix problems detected later, paid each time. I do not have that gift.

My points:

Subject is GEC's not EGC's. Multiple GEC's terminating on a common electrode busbar with steady state current always on. IMO if it was avoidable, then it's objectionable.

I am not proposing any added cost or voodoo methods like chemical rods or completely isolated earth electrodes, which do appear in specs.

Noise on the EGC can be expected and persist. There may be nothing that can be done. When the EGC is on the primary side of the transformer, each secondary N-G-GEC point gives you the opportunity to arrange, provide, a GEC that is relatively cleaner than the EGC. They connect, but they each have preferred paths for any current, noise, that may be on them.

The GEC implementation may be done compliantly with no added cost, but measurable improved value, by wisely choosing implementation method, path.

But by all means, if voodoo methods result in additional paying work, build a few more altars.

 

[kwired]

I am thinking there may be a way to put in ground rods at each location and then the run to the facility GES system is an electrode bonding jumper and not a GEC. The rods by themselves would do nothing but the savings may be in a smaller and less restricted run of wire. The electrode bonding jumper may be an easier less costly install compared to the equivalent GEC.

I have not checked this with the codes references I would need if I wanted to justify this. It is something I would consider looking at.

You can drive all the ground rods you want that is not a code violation, but you must still bond the grounded conductors of each system together so all systems have equipotential between their grounded conductors and equipotential will exist between non current carrying components that are bonded to the equipment grounding conductors of each system as well - including the service or other main building supply system. How it gets done may depend on permissions in code.

 

[Buck Parrish]

Okay , thanks. I can do that

 

[don_resqcapt19]

It becomes a fact either by the contract language, ie, whose discretionary judgement determines satisfactory work, or in the opinion of the enforcement official. Whose opinion it is may make it a fact.


It happens all the time. Some sensors work by generating very subtle electrical effects that are amplified and read by instrumentation.

I have seen strain gauges on steel beams, a lab test setup, that would not work. I found neutral current mixing on the grounding path. An audio recording studio will have extensive requirements to avoid electrical noise on the grounds (to provide a clean ground system). There are many many control and communication systems that can pick up ambient EMI in a way that is bad for normal operation. There are many occurrences of regular or persistent flow on the EGC's and other ground paths. It is routine.

Correctly designed EGC and GEC paths can mitigate this. Incorrect implementations can compound this.

Most of the time the "noise" is from the neutral current mixing on the ground paths....if that is happening it really has nothing to do with the EGC or GEC...it only has to do with illegal multiple neutral to grounding connections.

 

[__dan]

Most of the time the "noise" is from the neutral current mixing on the ground paths....if that is happening it really has nothing to do with the EGC or GEC...it only has to do with illegal multiple neutral to grounding connections.

Yes, but it happens with compliant systems also. You can expect it with dual fed bus if the bus is Y. You can unexpectedly, maybe legally, end up with more than one N to G system bonding jumper. One each at the multiple service transformers, more at the first multiple disconnects. N-G jumper locations can be hundreds of feet apart and you can end up with the neutrals and ground in parallel.

IMO, the way to do that cleanly is the dual fed bus would be 3 ph 3 w only, with only delta connected loads. No line to neutral loads on the dual fed bus, so there would be no neutral current to mix with the otherwise haphazard N to G bonding requirements. It's obviously not always done that way, so you can get a compliant noisy grounding arrangement.

Reactive power flow on large UPS on the ground paths is a big one. I don't know if the current generation new equipment is better. IMO, they would have to bring a line side neutral to it, or delta line side with a dedicated bonded path encircling the modules, would look good to me.

Everything can be expected to have some leakage current. It is not predictable that there will be none. Defective equipment, good equipment that is sensitive to or unreliable in the presence of EMI, the customer will want to run it all.

 

[don_resqcapt19]

Yes, but it happens with compliant systems also. You can expect it with dual fed bus if the bus is Y. You can unexpectedly, maybe legally, end up with more than one N to G system bonding jumper. One each at the multiple service transformers, more at the first multiple disconnects. N-G jumper locations can be hundreds of feet apart and you can end up with the neutrals and ground in parallel.

...

Now you are talking about service equipment....multiple connections from neutral to ground are required, and there is no way around that. You also have the connection on the utility side to their MGN.

With SDS, you should only have one point of connection.

 

[don_resqcapt19]

...
IMO, the way to do that cleanly is the dual fed bus would be 3 ph 3 w only, with only delta connected loads. No line to neutral loads on the dual fed bus, so there would be no neutral current to mix with the otherwise haphazard N to G bonding requirements. It's obviously not always done that way, so you can get a compliant noisy grounding arrangement.
...

If you make it an ungrounded system, yes, but that has its own associated problems. If you make it a grounded system, it is really no different from any other grounded system as there are required connections between the grounded conductor and the grounding conductor.

 

[__dan]

If you want to argue that noise does not happen, well it does. Noise can be expected.

The N-G-GEC connection at the secondary of a transformer is an opportunity to begin again with a system that can be relatively cleaner or away from noise that may exist on the primary side EGC grounding bus. None or very little primary side noise needs to flow over the *secondary side GEC*. Either can be arranged compliantly, a GEC that has flow and noise or one with much less.

The code provides the opportunity to remake the secondary side system cleaner. Implementation can achieve either. It is surveying, not ghost busting.

 

[__dan]

Go into the data center and randomly amp clamp the GEC on inside the building step down transformers. Especially the transformers also located near the big UPS but not part of it. You find 2, 3, 4 amps on the GEC.

Do you think it is 4 amps flowing to the earth as a source? It is not. It is a tiny proportion of the much larger flow on the entire grounded structure that the GEC happens to be intercepting and giving a shared parallel path to. Essentially a shunted current flow.

The GEC can be arranged to avoid this flow. The implementation should have this in mind.

 

[tx2step]

The GEC can be arranged to avoid this flow. The implementation should have this in mind.

I'm sorry, you may well be right, but this is nebulous to me. Would you provide "several" code compliant examples of bad implementation & resulting problems, along with code compliant implementations that solve those problems? I'd like to understand!

Thanks!

 

[__dan]

I'm sorry, you may well be right, but this is nebulous to me. Would you provide "several" code compliant examples of bad implementation & resulting problems, along with code compliant implementations that solve those problems? I'd like to understand!

Thanks!

I gave examples above where noise on the ground paths can be expected. If you survey and look for this, form an opinion of what may be causing it, a way to mitigate or avoid may become visible. First step would be routinely amp clamping the GEC's and occasionally, EGC's, looking for flow.

Most common is non compliance issues. No GEC of any kind is very common. Less common but found occasionally is no system bonding jumper.

Old commercial districts and large industrial buildings are frequently built up of several adjoining buildings added over time. They may look like one building from the outside but internally are several separate large structures.

Ideally there would be bonding jumpers from steel to steel but that's frequently omitted. The larger form can have many separate services and isolated power sources with the grounding and bonding occurring haphazardly, by accidental contact and not by intentional means, jumpers and runs of wire.

First step would be to survey with an amp clamp and see how clean or noisy the grounding paths are. The most commonly found problem will be the non compliance issues.

 

[iwire]

Dan, I have to be honest when I hear people say 'noise' to describe electrical problems I get very suspicious as many times these folks are making stuff up.

I am not saying you are, just that has been my experiance, it's like they are saying ghosts are involved.

My experiance has been more often then not that the equipment being supplied is faulty. Like the machine engineers that insist their CNC macine must be grounded via its own ground rod etc due to 'noise'.

For the past 20 years I have worked in the service end of large ECs that install countless transformers of various sizes each year and I just don't see them causing the troubles you are describing.

The OP is talking about a very basic very common installation.

 

[kwired]

Dan, I have to be honest when I hear people say 'noise' to describe electrical problems I get very suspicious as many times these folks are making stuff up.

I am not saying you are, just that has been my experiance, it's like they are saying ghosts are involved.

My experiance has been more often then not that the equipment being supplied is faulty. Like the machine engineers that insist their CNC macine must be grounded via its own ground rod etc due to 'noise'.

For the past 20 years I have worked in the service end of large ECs that install countless transformers of various sizes each year and I just don't see them causing the troubles you are describing.

The OP is talking about a very basic very common installation.

Just the other day I had a generically so called noise problem. I say generically because it is easier for the tech or engineer to say noise then to explain in detail what is actually going on, or in some cases they don't really know what is going on.

This was a radio controlled relay unit that allowed the operator to press a button on a wireless remote control to operate a control relay connected to whatever one desires to control with it, it had multiple relays and multiple corresponding buttons on the remote control.

We just recently connected this device and had troubles with the radio controller "locking up" almost right away, remove power to controller and it would reset but usually didn't take long before it would lock up again. When it locked up it wouldn't do anything outputs that were on stayed on and those that were off stayed off.

I called support and one of their engineers told me it was something they do encounter sometimes and is caused by inductive kickback from the load being controlled (which was simple "ice cube relays") but to talk to someone that don't understand what that is it may be just called "noise".

Anyway his solution was to place a capacitor across the relay coil, just did that yesterday and haven't had enough time yet to see if it completely solved the issue, but so far no trouble.

The funny thing that sort of relates to this discussion is he also mentioned lifting the EGC connection from the board on that controller might also solve the problem we were having.

 

[don_resqcapt19]

I still stand by my statement that most of the issues are caused by illegal grounded to grounding connections, and the guys on here who investigate these types of issues all the time say the same thing.

Yes, you can get a reduction in noise by installing a SDS, as you said you start over when you do that, but it is a very specialized installation where you have to take steps beyond what the code requires.

Another issue is the installation where the manufacturer's instructions calling for an non-compliant isolated ground have been used (an independent ground not connected to the electrical grounding system at any point) A code compliant isolated ground would be connected to the electrical grounding system at the location of the main main bonding jumper if the circuit does not originate at an SDS or at the location of the system bonding jumper if the circuit originates at an SDS.

 

[__dan]

My experience has been more often then not that the equipment being supplied is faulty. Like the machine engineers that insist their CNC machine must be grounded via its own ground rod etc due to 'noise'.

The OP's inquiry was addressed. For some reason the thread drew this 'noise does not happen' thread.

At the same time saying "noise does not happen", then saying well it happens, but only for defective equipment or non compliant installs, you are really admitting the falsity of the statement "noise does not happen". Noise happens and the code provides remedies to reduce or avoid it. A complaint installation may or may not be a remedy. Both happen.

The vast majority of issues or complaints I would say are due to defective faulty equipment and non compliant installs, but then you may be in the situation where you cannot fix (the customer does not want to pay for or give permission to fix) the problem and you must arrange your install to avoid or be immune to the present ambient noise on the grounding system.

I do not see where the argument is. Does noise happen, yes. If you don't want to look at it, fine. If you do want to look at it, go around and amp clamp all the GEC's and EGC's you come across in your daily routine and you will see some.

When you want some mitigation strategies, there will be an interesting area of discussion. Survey and run some wire from point A to point B. Determining A and B will either give an effective and compliant install or will not.