Interference from EMI radiation from service feeders?

[djr3203]

I have two sets of (8)4"C with (4) 600kcmil 3 phase 480V feeders running in a concrete encased duct bank supplying a switchgear from (2) 2000KVA transformers. The duct bank runs approx. 10 feet below a some rooms where there will be sensitive equipment for a bioengineering facility. The rooms will be constructed out of C.I.P concrete walls and floors. The client has expressed concern with interference from the duct bank.

Question:

1. Has anyone ever experienced trouble with EMI radiation from service feeders carrying this magnitude of current? If so, what was the distance from the feeders where this trouble was encountered?

2. Would the rebar reinforcement in the walls and floors of the room provide any protection from interference?

3. Does anyone know how to model the magnitude of EMI radiation off 3 phase parallel conductors?

The client is pushing for twisted feeders. My research has lead me to believe that the twisted feeders are primarily for ease in pulling the wires. I know twisted pair wires are used for signal wires to reduce common mode noise when transmitting differential signals.

Questions:

A. Would twisting the 3 phase feeders reduce the EMI radiated from the feeders?

I think the only way to truly cancel the EMI radiation would be to create an electromagnetic shield around the rooms or around the duct bank.

Questions:

B. I know an electromagnetic shield/faraday cage will cancel exterior noise from reaching the interior. Will an electromagnetic shield work in the other way and cancel the EMI radiated from the feeders outside of the shield/cage?

C. Could a wire mesh be cast in the concrete in a cage like structure around the duct bank and grounded to produce an electromagnetic shield?

Thank you for bearing with this extremely long thread. Please keep in mind that my electromagnetic theory fundamentals are quite rusty

 

[Skogsgurra]

I usually do not care to listen to prople that express concerns about radiation from cables. There is a subculture with all sorts of concerns and I sometimes think that they are concerned only because they 'have the right to be concerned'.

In this case, however, there may be a reason. I have no idea what EMI levels those bioengineering activities need. That is something the user shall define. If necessery, there are ways to reduce magnetic fields and electric fields. Electric fields are easy, magnetic not so easy. Some use heavy aluminum plates that are soldered into a closed compartment so that eddy currents shall counteract the external magnetic fields. And other use true magnetig shielding with high-permeability material.

A general concern is not good enough. You need data. And verification.

Rebar doesn't do much to the magnetic field. It just redistributes it somewhat.

Shields are symmetrical and are equally good at keeping fields out as they are keeping fields in - so to speak. A duct formed from mild steel and where you put the cables will probably work. Make sure it is closed with no discontinuities where magnetic field will concentrate and spread. There will probably not be any HF pollution in those feeders, so there should not be any concern about slot antenna effects.

Sorry. This was equally long. The main thing is to find out what levels of EMI that bioengineering facility can tolerate. Then perhaps discuss if they can move it somewhere else or if they are prepared to pay for investigation and mitigation.



Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[jghrist]

The only EMI that might be of concern is magnetic interference with electronic equipment. If the cables are in 3Ø groups and the loads are fairly balanced, the interference will be minimal because the phases will cancel.

Twisting will do nothing of any significance. Magnetic field depends on the distance from the conductor. Twisting would equalize the distance from the phases, but really, how much relative difference is there from phases once you are more than a few feet from the trefoil cables?

Use steel conduit if it will ease the client's concerns and if he will pay for it. This will put a Faraday cage around the cables. Steel conduit will increase the impedance of the circuit, so if voltage drop is a problem, it needs to be checked.

 

[waross]

If your client will be using a research grade Nuclear Magnetic Resonance machine, you may have to move either the cables or the lab.
Research NMRs are incredibly sensitive to magnetic fields.
Anecdote: I once took a friend of mine to visit another friend who was responsible for a large NMR in a university. I was showing my friend the intensity of the magnetic field by holding my key ring about four feet away from the magnet. The magnetic keys in the bunch were pulled out towards horizontal by the magnetic field. That's when my other friend started to holler at me. It was evident from the 'scope that something was distorting the magnetic field slightly.
I was then told about another research scope that was installed on the sixth floor of a building and suffered magnetic interference from subway trains passing underground beneath the building.
If your client intends to use a research grade NMR machine, something will probably have to move.

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

 

[djr3203]

Thanks for the info everyone. You are correct in the fact that I need to obtain more stringent criteria on what is considered "acceptable" interference.

jghrist, I am having difficulties understanding your statement " If the cables are in 3Ø groups and the loads are fairly balanced, the interference will be minimal because the phases will cancel." Can you provide a bit more insight? If they are 120 degrees out of phase with each other and the field is proportional to the magnitude of the current and the rate of change of the current, would they really cancel? I could see how the fields would cancel if you had 2 phases 180 degrees out of phase with each other.

Does anyone have any good links on this kind of info?

Does twisting reduce the EMI radiated from the conductors themselves? I am under the impression it only reduces noise produce on the conductors from outside interference only.

David

 

[rcw retired EE]

For any 3-phase, 3-wire system, at any instant Ia+Ib+Ic = 0, unless a fault occurs. If the currents sum to zero, the magnetic fields generated by those currents also sum to zero, if the measuring point is far enough from the three conductors to minimize the error caused by conductor spacing.

Placing each cable group (A,B,C,G) in a steel conduit for the whole length will attenuate the field a lot more than twisting the phase wires.

Your circuit will be about 4,000A max load. We had problems with magnetic fields affecting CRT's when a 4,000A non-segregated phase bus duct ran on the electrical room ceiling just under the plant manager’s 2nd floor office floor. The asymmetry of the bus bars created large magnetic fields that twisted the CRT picture. Our construction group placed 0.25" steel plates on the floor but did not fully weld the seams. The plates mitigated the field in the middle but concentrated the fields at the joints.

Your cable in conduits should not have this problem. As suggested above, add a bonded layer of mu-metal on the floor.

 

[djr3203]

Stupid me I know the currents sum to zero on the neutral for a balanced load, which obviously means the summation of the fields for the three should sum to zero.

So if the symmetry of the cables are equal and they carry the same current, the fields will sum to zero. This means that in a normal situation the only EMFs radiated by transmission would be any harmonics on the line and/or if there is an unbalanced load?

So rcwilson, in your case it was the actually geometry of the bus bars that were different from phase to phase which resulted in the fields not canceling?

Does a non-segregated phase bus duct just mean that all the bars for the different phases are run in the same enclosure?

 

[jghrist]

The summation of the fields for the three phases will only sum to zero if the distance from each phase is equal. This is nearly true for cables installed with all three phases in one conduit. Twisting the cables would make the average distance to each phase the same, but I wouldn't want to be the contractor trying to twist 600 kcmil cable while it was being pulled.

 

[rcw retired EE]

djr32003- Non-seg bus duct has the phase bus bars side by side in a common metallic enclosure. The 13.8 kV, 4000A bus had two 1/2" x 8" insulated copper bars per phase oriented vertically with 7.5" phase-to-phase spacing inside a 20" x 30" aluminum enclosure. The bus duct was hung just under the 2nd floor support steel so the top of concrete on the second floor was only 20"-24" above the bus centerline. The manager's CRT was about 3-1/2 feet above the bus. The A-C phase spacing of 15" was around 40% of the CRT to bus distance, not far enough away for the fields to cancel.

The screen images would move and jump as the generator load changed.

As a side note, the client's manager who occupied that office was extremely concerned about magnetic fields from his previous career as a Navy submarine officer responsible for the sub's generators and electrical systems. He claimed his medical records proved his white blood cell count dropped significantly during the times he worked near the generator's magnetic field whether he was at sea or doing research in some lab.

The karma was we did not design any office space in the area above the bus duct. That manager changed the building and room arrangement to get an office with a 2nd floor view. He didn't realize until it was built that the GSU blast wall was 3 feet outside his window or that he sat directly above the bus.

In your case, the conductor spacing is small in relation to the distance to the room and steel conduit should minimize the external fields. There still may be some fringe effects that may require adding some mu-metal barriers.

 

[djr3203]

Hey thanks all for the post and the info. All the post have been helpful.