Grounding an ungrounded delta

[DaveMatlock]

We recently had a costly incident in our mill that resulted in the destruction of 6 VFDs. These drives are in a section of the mill that is supplied by a transformer with a ungrounded 480 V delta secondary. It appears that the damage resulted from over voltages created by an intermittent arcing fault in an underground feeder.

In order to prevent future damage from similar incidents, I believe that we need to get the system grounded. My first inclination is to simply corner ground the transformer and monitor for ground current so that we know if the B phase gets grounded elsewhere. However, I am also considering using a Zig-zag transformer to create a neutral point.

Outside of the issue of resistance grounding, are there any advantages of the Zig-zag transformer over a corner grounded system? Are there any unforeseen problems that could result from corner grounding?

 

[dpc]

I completely support converting the ungrounded delta to a grounded system. While the corner grounding can probably cure the repetitive restrike overvoltages, I would recommend a grounded neutral system.

As a minimum, you will want to check with the VFD manufacturers, since I believe most VFD front-ends expect a grounded neutral supply.

I don't have a lot of experience with corner-grounded delta systems - and neither will anyone else at the plant, I suspect. It will introduce a lot a new ways to make mistakes and bad assumptions by anyone not familiar with your system.

A grounded neutral system will be familiar to a journeyman electrician and will be much easier to maintain and troubleshoot.

As for grounding transformer, a standard delta-wye transformer can also be used as grounding transformer - a zig-zag unit is not required. If you go with neutral grounding, I'd recommend high-resistance grounding. But again, verify with VFD supplier that this will not be an issue with the drive.

David Castor

http://www.cvoes.com

 

[rbulsara]

One of the considerations would be that many VFDs like balanced wye- three phase source, even if no neutral is used. Their components are designed for line to ground voltages. Zig-zag grounding will provide that wye system. There are members here with expertize in VFDs who can shed more light on this.

Beside that there is no real advantage that I know of. But a balanced system with a neutral would be more familiar to operators and electrician, who are used to a grounded wye systems, from maintenance and testing points of view.

Corner grounding is perhaps cost effective but requires a little more care in maintenance and trouble shooting. It is also becoming very uncommon.



Rafiq Bulsara

http://www.srengineersct.com

 

[Borti]

You can check this site

http://www.ipc-resistors.com/default.asp.

They have complete solutions.

 

[DaveMatlock]

Thanks for the posts they are very helpful.

I would prefer to go with a grounded neutral setup but I thought that corner grounding would be a quick easy and inexpensive way to solve the problem. Also, none of the electrical suppliers that we use had ever heard of a zig-zag transformer, so I was not sure where to get one.

Dpc, thanks for pointing out that a standard delta-wye transformer would work in place of a zig-zag transformer. I assume that with high resistance grounding that you would limit the ground fault current to a few amps and therefore you would not need a very large transformer. Is this correct?

 

[dpc]

Yes, the transformer kVA requirements are pretty small. There are several suppliers who can provide the grounding transformer and grounding resistor as a package, which is probably the simplest approach at 480 V. In addition to IPC already mentioned, check Post-Glover as well as GE, Cutler-Hammer and Square D.



David Castor

http://www.cvoes.com

 

[SparksRfun]

Corner grounded delta systems are common in old industrial facilities around St. Louis, and they are a nightmare. My recommendation on corner grounded delta is to bolt all three phases together and hire a monkey to throw the main switch!

But seriously, you'll have nightmares with corner grounded delta, it causes wierd ground currents to flow in unexpected ways. If we renovate anything with corner grounded delta, the power company will not connect to it again until we fix it. sometimes we've had to put in huge 480Y:480Delta transformers before the power company would even touch us.

Establish a neutral that's your best bet.

SparksRfun

"The truth will set you free, but first it will make you madder than a wet bobcat"

 

[davidbeach]

Instead of a zig-zag you can use a wye-delta transformer. You need to have a 480V wye, and ground the neutral. The delta side doesn't matter as it doesn't need to be connected. Put a breaker on the transformer, add an aux contact so that you can trip the main when the breaker on the ground transformer trips, and you're good to go.

 

[odlanor]

Another problem is the representation of system for short-circuits calculation.
With neutral grounding(grounding transformer or Y_D transformer) you can use symmetrical components for calculation
and you will have problem with corner grounding.

 

[waross]

If you have a transformer with a center tapped 480V winding, you may connect that in parallel with one phase and use a four wire grounded delta system. You will have two phases at 240V to ground and one wild leg at about 416V to ground.
Poor man's solution.
I prefer davidbeach's suggestion. Wye delta transformer.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[busbar]

Understand that using a Y-? transformer, if the wyepoint is grounded through a resistor, and the ? left floating, then you have a high-resistance-grounded system. A good starting point for offsetting charging current is to size the resistor at 1 amp/MVA of transformer capacity.

 

[DaveMatlock]

Thanks to all who posted for the information provided. I am currently weighing my options.

While I do like the idea of the wye-delta transformer suggested by davidbeach and others, I have also looked into the grounding transformer and resistor package units from several manufacturers. They all are pushing their own version of pulser units. These units are supposed to aid in locating ground faults by creating a higher current pulse that can be seen on an amp meter. The basic idea behind them seems simple and I have no doubt that they work as advertised, however, I am not sure that the procedure used to locate the fault it practical.

Have any of you had any experience with these systems? Are they worth the added expense?

 

[dpc]

The pulsers generally work as advertised, but it's still going to be a bit of trial and error to half-split the system and locate the fault. The process does work, however and it is much faster with the pulser than without. You can buy the pulsers as a standalone product also.

But I wouldn't consider a high-resistance grounded 480 V system without some type of fault location device.

David Castor

http://www.cvoes.com

 

[busbar]

A good IEEE paper on the practical aspects of high-resistance grounding is

http://neiengineering.com/pdfs/paper1JN.pdf