Grounded vs ungrounded service

[PipelineJim]

I was recently brought in on a situation wherein our incoming service feed (480vac 3 phase ungrounded) experienced a failure. A discussion quickly ensued about whether or not we should/must replace this with a grounded three phase service. I felt that we are required to do so, simply for safety reasons if no other. I was asked to explain why a three phase 480vac service should be grounded, or if it could remain ungrounded (floating). My answer was that a transformer is grounded to provide a reference to ground for overcurrent/fault protection as well as to balance the incoming power. I was then asked to explain exactly how that works and I was at a bit of a loss for words. I have tried to find answers to this online, but have not been able to find any specific information about why three phase transformers should be grounded as opposed to ungrounded, and the specifics related to the subject.

My experience with ungrounded services is a bit limited, as I have worked on grounded systems for most of my career and have only seen a few ungrounded services.

The second part of this question is if the facility equipment (switchgear, devices, etc.) are grounded to an existing grounding grid, does the incoming service have to be grounded at the purchase-power transformer? and why?

Any information I can get on this will be greatly helpful and appreciated.

 

[augie47]

"ungrounded" services are legal and are a choice. There is no requirement to connect to the premises ground grid unless your power source has a grounded conductor.
The primary advantage is the ability to continue operation of equipment even when 1 phase goes to ground" allowing you to trouble shoot the failure while still operating.
NEC 250.21(B) requires ground fault detectors on these systems.
The disadvantage is the possible potential between poorly grounded items, and the cost in not being able to use slash-rated breakers..

I would think most electricians would consider a grounded system to be a bit safer.

 

[PipelineJim]

Would I also be correct in thinking that our fault protection (IE: fuses) on the incoming service would only protect against phase-to-phase faults and not phase-to-ground faults? Speaking about an ungrounded service of course. A grounded service would provide both types of fault protection...if I am thinking about it correctly.

Thank you again for the insight. I have been under the impression that ungrounded systems were not allowed by the NEC. I am very glad to get some clarification on the subject.

 

[augie47]

you are correct.. that's the advantage... a phase to ground short will not open the OCP device. with the indicators, you will be alerted to the "fault" and can take corrective action.
If there is a phase to ground fault on a 2nd phase prior to your correcting the 1st fault, then you will have operation of your OCP (occasionaly dramatic )

 

[jim dungar]

Would I also be correct in thinking that our fault protection (IE: fuses) on the incoming service would only protect against phase-to-phase faults and not phase-to-ground faults? Speaking about an ungrounded service of course. A grounded service would provide both types of fault protection...if I am thinking about it correctly.

Thank you again for the insight. I have been under the impression that ungrounded systems were not allowed by the NEC. I am very glad to get some clarification on the subject.

Separately derived ungrounded systems are only allowed in industrial facilities where supervision assures 'trained/qualified' persons perform the maintenance, 250,21(A)3. By you and others at your facility asking these questions I wonder if an ungrounded system is still appropriate for you. I am glad you are asking questions on this forum.

Prior to 2005, ground detection/alarming was a FPN, it is now required, 250.21(B).

The first phase-ground connection is not truly a fault and so would not need to be cleared by your fuses. However, it should be discovered and removed as soon as an orderly process shutdown can be arranged. If the first connection is an arcing one, transient overvoltages may be experienced and unsafe touch potentials to other grounded surfaces, in the facility, may be possible.

 

[PipelineJim]

Speaking for myself, I do not like ungrounded systems for exactly the reasons you are giving. I want a fault to be cleared by a fuse or other fault protective device as soon as it occurs.

The questions were being asked of me from a technical standpoint as to the "hows" and "why's" pertaining to how a grounded transformer handles faults in comparison to the ungrounded system. We are looking at a very large expense if we are required to go to a grounded system because this service was installed in 1970 and the utility is uncertain(as we are being told) if the transformers can be grounded or if we must purchase a new transformer in addition to the new conductors we must pull. Because of this, they want exact answers as to how the ground provides a reference to ground and how that affects the fault/overcurrent protection of the transformer and how it affects the equipment inside the facility that is on a ground grid and if the NEC requires us to make it a grounded service to begin with. So, while I understand grounded transformers and fault protection, I was having some serious mental blocks when it came to explaining the ungrounded system.

For example, when I stated that the transformer needed a reference to ground to provide fault protection (via the transformer fuses) in case of a phase-to-ground fault; I was asked why the grounding grid inside the station wouldn't provide that for the transformer. When I said that the transformer needed it's own reference to ground, I was asked.....why? That's when the mental blocks started. Anyone who has had to try to justify large, emergency expenses like this knows the types of questions that are asked; and although we tend to know the answers, it is difficult at times to pull them out of the closet of antiquity. I explained about phase-to-ground on an ungrounded system and the overvoltage situations it can create, as well as the affects that it can have on equipment; but what they were looking for is an answer to why they should spend the money to make it a grounded system when the ungrounded system has worked for the past 40 years.

I am very glad to have this forum as a resource to get some input. Thank you for the assistance.

 

[SG-1]

This link compares different grounding systems.

http://www.postglover.com/TechLibrary/Grounding_Methods.pdf

It is possible to derive a neutral using several different methods.

 

[augie47]

If your supply is ungrounded and you don't like an ungrounded system and you like a "reference to ground", then ground 1 phase and have a corner grounded system. (you will need to address any fuses on the grounded system)

 

[tom baker]

Another option is the high impedance grounded neutral system. It combines the fault clearing and stable voltage of a grounded neutral system with the fault tolerance of an undgrounded system. A resistor is used as part of the main bonding jumper, on a ground fault the resistor develops a voltage drop, sounds an alarm, to allow the system to be shut down in a timely fashon.