HV Lab Grounding 1

[VTer]

Hi gentlemen,

I have a High Voltage Impulse generator that will be used for testing. We are dedicating a room for this equipment and I have some concerns. The manufacturer recommends isolating the test room grounding system from the rest of the building grounding system. They are proposing a PVC barrier between the test room concrete floor and rest of the building floor. They are then recommending a grounding grid and a single rod driven down enough to provide a ground resistance less than 2 ohms. They are also recommending that even the steel columns and structure in the room be isolated from this isolated floor. My concern is step potentials between the two isolated floor slabs and also touch potentials from isolated steel structures. Do you guys see any problems with this recommendation? Are my concerns valid? My thoughts are to provide ground rods as recommended to get the resistance less than 2 ohms but then I would bond all the steel in the room to this grounding system and also I would bond to building grounding system as well. Are there any concerns with this? I hope somebody has some experience with this, as this is my first HV lab. By the way, the generator is 500kV with 5-100kV stages if that is of any relevance to you guys. Any recommendations will be appreciated. Thanks in advance.


"Throughout space there is energy. Is this energy static or kinetic! If static our hopes are in vain; if kinetic — and this we know it is, for certain — then it is a mere question of time when men will succeed in attaching their machinery to the very wheelwork of nature". – Nikola Tesla

 

[alehman]

I don't know if anyone here has experience with such things. I do not. However I would share your concerns. In general I'm not a fan of isolated grounding systems because they are difficult to accomplish in a practical sense and safety is always a concern unless carefully designed, installed and maintained.

Alan
----
"It’s always fun to do the impossible." - Walt Disney

 

[rbulsara]

I was also waiting for someone with such experience to chime in. I am also not experienced in 500kV test labs but I cannot endorse separate grounds as it is against all codes, standards and practice for personnel safety.

I am not sure what use the 2 ohms ground would be. If this is a test lab, it should have its a solid metal cage around it and some solid base of an array of steel framing to which all of its references, neutrals or grounds should be bonded to. And that frame should be bonded to rest of the building's steel and grounding electrode. No current should be going to the actual earth in a test environment. Steel ground frame should take care of any fault or test discharge currents paths back to the source.

I would be more than happy to red flag my own post should someone with extra HV experience chimes in here with a different view.



Rafiq Bulsara

http://www.srengineersct.com

 

[xxjohnh]

You could design it like an isolated instrument ground system. Insulate it and have a removable link for a single point connection to the building ground. The link is normally closed but could be opened for denenergised testing.
You could make a drawing and send it to the vendor for comment, then submit it, if required to any other approval body.

 

[rbulsara]

xxjohnh:
What is "like an isolated instrument grounnd"? Sounds like a code violation.

Rafiq Bulsara

http://www.srengineersct.com

 

[xxjohnh]

Distributed Control System vendors have for years required an isolated earthing system sometimes with a dedicated earth electrode, others we just single point connection to the main grid.Most code authorities accept engineered solutions for special appications, such as this vendor is requesting.

 

[VTer]

Hi guys, thanks for the replies. I have been doing much research and there is not much out there on this. IEEE Std 4-1995 does not say much if anything about grounding, and any other references I found regarding grounding do not match with each other. The little suggestions and recommendations I found are all over the place from completely isolating the room to completely enclosing a room in a metal frame to single point grounding. I hoped someone on this helpful site would point me in the right direction. As I stated in OP I feel most comfortable with bonding to building grounding system and by grounding everything in the room. We will also provide a small 10kVA transformer to feed the charging rectifier in which case I think grounding rods would be required as that would be considered a separately derived system. BTW I think the 2 ohm recommendation is for test accuracy and protection of sensitive electronics which are a part of this HV system.

"Throughout space there is energy. Is this energy static or kinetic! If static our hopes are in vain; if kinetic — and this we know it is, for certain — then it is a mere question of time when men will succeed in attaching their machinery to the very wheelwork of nature". – Nikola Tesla

 

[rbulsara]

xxjohn:
That is a clear code violation in the USA and also not in conformance with IEEE standards, Check out the Green book and the Emerald book, plus NEC. You have been listening to wrong vendors.

I know of no such authorities accepting those.

Rafiq Bulsara

http://www.srengineersct.com

 

[xxjohnh]

Read some ABB, Yokogawa, Honeywell or Taylor installation manuals. It is very common to have isolated instrument grounding systems.

 

[magoo2]

Might be useful to ask PowerTech Labs or Kinetrics and see what they do.

 

[jghrist]

Isolated grounding is allowable, but it is not completely isolated. There is still a connection to the main grounding electrode at the source.

See Demystifying Isolated Grounding Systems — Parts 1 & 2

http://login.ecmweb.com/wall.aspx?E...ic_demystifying_isolated_grounding/index.html

and

http://ecmweb.com/mag/electric_demystifying_isolated_grounded/index.html

 

[dpmac]

I posed this question to one of our engineers who is involved with large test labs all over the world. His response is below:

Most manufacturers of high voltage test equipment recommend a fully isolated Area, with one, two or three grounding rods paralleled to achieve 2 ohms or less.

The main reason for this recommendation is to have full control of the grounding during installation and commissioning of their equipment. The transients generated by an impulse generator are similar to lightning strikes and high voltage switching disturbances, both of which commonly cause problems in a normally grounded facility. It is normally possible to have a single point connection between the test area ground and the building ground, but the exact position of that connection may not be able to be determined until the equipment is installed and tested. It is much easier to make connections to building ground after the equipment is installed than it is to break connections.

There is no building code in the world which encourages such a scheme, although there are probably hundreds of such test environments in existence. The chance of touch potential or step potential should be non existent, as a high voltage test area should be surrounded by a 6 or 8 foot high safety fence, bonded to the factory ground. This fence is primarily designed to prevent injury from the high voltage test equipment, but it can be argued that it is also their to prevent any contact with the isolated area whilst the equipment is energized.

In reality, depending on the energy per stage of the impulse generator 500kV is not a particularly large test system and it would probably not matter if everything is bonded together in the beginning, but I have personally seen a whole office full of computers stop working when firing a 2000kV generator due to a facility not following the manufacturers grounding recommendations.

It is also worth noting one other thing. Most manufacturers of high voltage test equipment will not honor the equipment warranty, or any damage caused to their equipment, or to any sensitive equipment within your facility, unless you followed their recommendations with respect to test area layout.

 

[scottf]

I agree with the comments on dpmac's post.

We're in the process of building a new factory and the grounding of the test equipment is a big concern.

As mentioned, transients as a result of improperly grounded impulse generators can be a big problem. In one of my former factories, we used to blow out expensive network switches all the time until we finally figured out we had a grounding issue.

Also, if partial discharge measurements will ever be performed, the ground system is even more important. It required the lab ground to be completely isolated from the building ground if doing sensitive measurements (which pretty much all PD measurements are). It's impossible to measure <2 pC when you're picking up noise from the factory side.

 

[mgtrp]

"It required the lab ground to be completely isolated from the building ground if doing sensitive measurements (which pretty much all PD measurements are)."

Agree with the interference issue - I could tell which other equipment was running in factory I used to test at by the "signature" on the oscilloscope while PD testing. Very frustrating having to wait for it whatever machine was running to stop!

 

[ausphil]

Grounding loops in test areas do cause major problems with PD (as mentioned) with high current applications (induced currents into ground loops make for incorrect measurements and unsafe voltages at points that you think are "grounded") and for impulse tests, where you can cause all sorts of power quality problems for computers and other electronic equipment from the spikes generated into the local power system ground system. Radial ground leads to equipment is the way to go to eliminate these things.

Also, the concept for ground return current when high voltage testing is to try to minimise the impedance when there is a failure, and control the fault current path. By having multiple paths, you don't know what is happening, and may acutally transfer lethal potentials outside you test area.

There is a book by Nils Hyleten-Cavallius called "High Voltage Laboratory Planning". It gives a pretty good rundown of all requirements on HV labs, including clearances, grounding, shielding etc. It is from 1986, but the concepts are still valid (even if the test and measurement equipment are outdated).

My suggestion is if you haven't been involved in a HV lab before, get people in that have designed and used them before.

If the manufacturer is a reputable one, then their suggestions are to be listened to. People that have setup labs before should know the local wiring requirements.

ausphil

 

[VTer]

Thank you for all your replies. The information you provided is much helpful. I have scheduled another phone conf with the manufacturer for tomorrow. Now I understand some of their reasons for isolation from building grounding system but I am still concerned with their recommendations not to bond building columns and steel in this room to this test grounding system. My first concern is human safety, the accuracy of test and equipment frying is a concern, but comes second. I will keep everyone updated. Thanks again.

"Throughout space there is energy. Is this energy static or kinetic! If static our hopes are in vain; if kinetic — and this we know it is, for certain — then it is a mere question of time when men will succeed in attaching their machinery to the very wheelwork of nature". – Nikola Tesla

 

[dpmac]

As our engineer mentioned in his post above:

"The main reason for this recommendation is to have full control of the grounding during installation and commissioning of their equipment. The transients generated by an impulse generator are similar to lightning strikes and high voltage switching disturbances, both of which commonly cause problems in a normally grounded facility. It is normally possible to have a single point connection between the test area ground and the building ground, but the exact position of that connection may not be able to be determined until the equipment is installed and tested. It is much easier to make connections to building ground after the equipment is installed than it is to break connections."

You may very well be able to provide additional and proper grounding AFTER the IG Set is commissioned.

 

[VTer]

Dpmac,

Thank you and your engineer with the helpful feed back. I do take serious consideration of your statements above, but I do however need to verify such information with the manufacturer. In any case I will keep everyone updated with their installation/ commissioning/ and operational approaches and concerns to see if they deviate from what has been discussed in this thread. Thanks again.


"Throughout space there is energy. Is this energy static or kinetic! If static our hopes are in vain; if kinetic — and this we know it is, for certain — then it is a mere question of time when men will succeed in attaching their machinery to the very wheelwork of nature". – Nikola Tesla