System and Equipment Grounding 1

[a10jp]

I have asked this question to many folks before. And I always believe solidly of the grounding principles in NEC. I recently found myself in doubt whether system grounding and equipment grounding should be connected together, which shows how poorly I understand this subject.

Recently we installed a new X-Ray machine in a clinic. This is a 400V unit supplied by a brand new 150kVA, 6.6kV-420V, 3 ph 4W, delta-wye transformer. The transformer is co-located together with 2 other 300kVA, 6.6kV-208/120V, 3ph 4W, and 150kVA 6.6kV-400V, 3ph 3W(center tap) within an enclosure (substation).

The equipment manufacturer insisted that X-Ray equipment be supplied with a separate grounding electrode dedicated for the X-Ray control cabinet, which is separate from the system grounidng at the substation. (At the substation, all the transformer shared a common ground.)

Although I mentioned that the separate X-Ray grounding electrodes (Equipment ground purpose) should be tied in with the service ground at the service disconnect (System ground purpose), per NEC 250-2, the equipment manufacturer voted against and directed only separate ground be used.

We are located in Japan, the grounding concept is different than in US.

I was explained by a local Japanese engineer that because we are dealing with different voltage system, the grounding at the substation should not be tied together and should be kept separate. And because System ground and Equipment ground serves entirely different purposes, they should remain separate. The reason is when there is a fault at one of the phases, the grounding conductor serving as a ground fault return path will carry a different potential and damage the equipment should they be connected at the source.

I would like to get some advice on this. I am very confused.

 

[Skogsgurra]

There should be an FAQ on this. Really.

MHO (as in IMHO) is that grounds (all grounds) shall be tied together (bonded) in one well defined point. Usually where the main entry point is. And that the earth rod(s)/mesh/lines shall be connected to that point also.

Then, I do not understand why the fact that System Ground and Equipment Ground are said to serve entirely different purposes should change this fundamental principle. I do not even understand what the "entirely different purposes" are. One ground for shields and one ground for protection? That doesn't change anything and that has been discussed over and over.

There are TT systems, where transformer is grounded "there" and equipment is grounded "here" and no bonding between the two. This is, as I see it, not a very good technique and should not be used in modern plants.

Many instrument and equipment company representatives have a very vague notion about ground, earth, bonding etcetera. They even sometimes confuse neutral and ground/earth. I do not know Japan Electric Code, but I would be surprised if it is any different from what is being done in other countries. Ask the local engineer to explain (draw a simple diagram) how the other equipment can be damaged if the two systems are solidly bonded together. Ask him to illustrate what happens when an earth fault causes overcurrent and different voltages in the two systems (grounded together and not). I think that will clear things up. And I guess that you will have it your way, too.

If it turns out (after all) that it shall be done "their way", please feed that information back to this forum. It would be interesting to know why "our way" doesn't work in this particular case.

Gunnar Englund

http://www.gke.org

 

[waross]

Hi Gunnar.
I agree with you completely.
When you consider the relative ease with which a simple analysis of a ground system can indicate the potential differences that can be developed by the passage of currents through the impedances of the components of the ground system,
When you consider the numerous examples of touch and step potentials developed by ground currents,
When you consider the magnitude of the earth voltage gradients that station engineers must consider in their designs,
and when you then consider some of the grounding schemes that the other disciplines pass down more as folklore than as anything that has been investigated and/or calculated, I think this may be more a topic for the ethics forum than for the Power forum.
I liked the grounding suggestion that you proposed in the last grounding thread, Gunnar.
a10jp
You may try to ask the vender for a ground safety analysis and a written undertaking that his grounding system will be safe in the event of a ground fault event on either the xray equipment or on nearby equipment grounded on the main system ground. Particular attention should be paid to possible voltage differences between nearby conductive surfaces that are connected to different ground systems. Attention may also be drawn to possible liability in the event of an injury or fatality in the event of a ground issue. The vender should accept this responsibility in writing.
Equipotential protection consists in part of:
1> Maintaining all exposed metalic and/or conductive parts at the same potential by interconnection with suitably sized conductors, and
2> Recognising that a current flow in an impedance develops a voltage and arranging the grounding connections to reduce and avoid as much as possible the flow of fault currents in conductors that would tend to produce a voltage difference between the exposed conductive parts, and
3> Recognising that ground fault currents from other sources may be thousands or tens of thousands of amps. These currents may fuse conductors and/or cause the failure of connections. A well designed ground system will pass such currents in such a way that any failure will not disrupt the
interconnections of the protected exposed surfaces so that the equipotential protection remains intact.
Respectfully

 

[a10jp]

For the case of system grounding alone, do you all recognize the same whether we have 400V system, 200V system, or even single phase system (1 ph, 3W), where all the grounding are connected together also? The Japanese engineer has made one example on this point to show show fault from one system cause large voltage potential to others if they are connected together. (Given all ground points resistance are relatively the eame...25 Ohms) At that point I was uncomfortable to assue myself which method is correct...

 

[a10jp]

I left out one important differences that I have forgot to mention. In the current service disconnect setup, the menufacturer require us to install ELCB breaker at the servuices disconnect, which is unusual for US design. But after some investugation, it appears this ELCB is used for ground fault detection since if you do not connect the separate equipment grounding electrodes to the system ground, there is no return ground fault path in the system, and so they install ELCB breaker for this purposes.

 

[ScottyUK]

a10,

That is starting to sound eerily like what we in Europe know as a TT system, with a local grounding electrode and an RCD (ELCB) at the service entry. They're a poor choice if you have option of a proper earth connection to the building main earth terminal.


----------------------------------

I don't suffer from insanity. I enjoy it...

 

[Skogsgurra]

????

"Given all ground points resistance are relatively the same...25 Ohms"

Have you been discussing ground currents and potential distribution assuming that different systems all have a 25 ohm ground resistance?

Has the Japanese engineer shown that there will be a potential difference between the systems if you have an earth fault in one of the systems and they are not bonded? Sure, there will be a potential difference. Ohm's law shows that convincingly.

But if you bond the two systems (and equipment in the systems), there will not be any potential difference. Especially if your bonding is made the correct way and with adequate copper area. That's what bonding is for.

Gunnar Englund

http://www.gke.org

 

[a10jp]

Hi Gunnar,

Here is what I take from the Japanese engineers, his argument is based on the case if all grounding for various voltage systems are connected togther. I used 25 Ohm as the worst case. Typically if you connected more copper electrodes over a decently large area, you can achieve less ohms than 25 Ohm. He said supposed you have a fault at the high voltage system, the fault current will travel through the grounding grid, which may be high for a lower voltage system connecting to the same gporund grid. Does that make sense? The 25 Ohm is an example he used, since it is common to assume all grounding system should be 25 Ohm or less. Then here is another scenario, say the system ground for a 6kV system has a fault, how would that affect a piece of sensitive equipment down range connecting to a 120V panel ground bus? That's the argument I was uncomfortable about, sharing grounding with totally different systems. And here is another point, if the Japanese are so adamant about their system, it might be becasue they have been working for so long, that they do not see any reason to change, and if that is true, are we saying both systems are correct and safe? But how could 2 systems be correct at the same time?

 

[waross]

Hello a10jp
If one system is grounded with 25 ohms resistance to ground and the other system is grounde with 25 ohms to ground the resistance from one system to the other is 50 ohms. It won't take much current to develop high voltages between the two ground systems.

The Japanese engineer has made one example on this point to show show fault from one system cause large voltage potential to others if they are connected together.

If the engineer can demonstrate this then he has not made the connection between the two systems properly.
When the grounds are properly installed, a fault from a different system will not cause a current between the two systems. No current, no voltage difference.
If your engineer removes the ground electrodes for the xray machine and leaves the ground systems interconnected and rechecks his calculations he will find that there should be no voltage difference between the systems. He is welcome to add the ground rods to the main ground system, they will do no harm there.
respectfully

 

[Skogsgurra]

I do not want to carry this discussion any further until I get a clear picture of what the alleged risk is and how a well bonded system would increase that risk.

Good practice, ohm's law, electrical codes of different countries and common sense are all facts that speak against the separate grounding scheme.

Show us an example with fault currents and give us the numbers used by your Japanese engineer. That may clarify what scenario he is trying to avoid by using separate grounds for system and equipment.

If you need to post a diagram showing the plant's ground system, see

http://www.eng-tips.com/faqs.cfm?fid=1161

for instructions how to do that.

Gunnar Englund

http://www.gke.org

 

[stevenal]

There should be an FAQ on this. Really.

Yes really. Wasn't the last time just last week? Looks like the number of FAQs skogsgurra has written to date is zero. Time to make it one? I'll promise to go easy on the rating.