Calculating EMI 1

[steven7]

Does anyone have an idea on how to calculate the EMI in an electrical room given the nameplate ratings of the gear: switchboards, branch panels, transformers, feeder sizes etc....

any help would be appreciated,

-Steve

 

[IRstuff]

Sorry, it doesn't work that way

TTFN

 

[Warpspeed]

Not possible steven.

 

[tinfoil]

Steven:

You will find that, even within a particular model of a particular mfr's equipment, there is significant variation in radiated EMI, due to tiny variations in construction (there's always some sort of tolerance), coupled with a very unpredictbale loading pattern in the equipment.

The only reliable way to get an EMI for a particular setup is to to empirical measurements. Even these will be suspect as emitted EMI is dependent on the specific loading conditions at the time of the testing.

 

[cuky2000]

The total EMF around the switchgear could be determined by a combination of measurement and calculation as follow.

[blue]Fe = (Fx² + Fy² + Fz² )^1/2[/blue]
_{Where,
Fe= Resultant (root-mean-square) sum of three internal orthogonal magnetic fluxes
Fx, Fy, Fz = are the magnetic flux densities obtained by the Gauss meter for the three directions coils on the x-, y- and z-axis,.}
See a sample of a Gauss meter in the enclose site:

http://www.omega.com/ppt/pptsc.asp?ref=HHG-10&Nav=f01

[blue]The enclose images are from VitaTech Eng.:[/blue]

 

[EEJaime]

Unfortunately the fellows are correct. It is virtually impossible to predict what the EMI levels of any installation will be. They are impacted by structure, feeder routing, panel loading, voltage levels, and any number of other factors. It just cost us about $22,000.00 to shield an office on the opposite side of an electrical room containing a switchboard and UPS unit because one of the community college professors got his hands on a Gauss meter and raised Cain telling the administration we were giving all his students cancer. It's insane, but if you are designing a new facility and you have large electrical equipment anywhere near occupyable space, I would recommend specifying a shielding system ahead of time. It will be much less expensive in the long run than the alternatives.

 

[cuky2000]

EEJaime,

I am under the impression that what did not work in the case described is the shielding of the room not the EMI measurement of the existing installation.

Shielding effectively a room is not an easy task and very expensive proposition. However, the level of EMF is possible to be determining with a reasonable degree of accuracy in existing facility.

 

[EEJaime]

I understand, but my impression is that Steven7 was looking for a method of calculating EMF before the equipment is installed. If I misread that-that is my mistake. In any case, I feel that all the studies I have seen, including white papers by some utilities have shown little correllation between EMI/EMF and health problems, although of course there is impact to sensitive electronics, etc....

Have a pleasant day all.

 

[itsmoked]

EEJaime: What was the basic shielding method that was used?
Could one not just put up IRON sheeting?

 

[jghrist]

The original poster asked about EMI, or Electromagnetic Interference. EMI is a combination of terms that broadly refers to any type of interference that can potentially disrupt, degrade or otherwise interfere with authorized electronic emissions over approved portions of the electromagnetic spectrum. (

http://www.chips.navy.mil/archives/01_fall/electromagnetic_interference.htm)

Radio Frequency Interference (RFI) is a form of EMI and is what normally is objectionable.

EMF, as described by cuky2000, is Electromagnetic Fields, in particular, magnetic fields or flux.

EMI is radiated energy and EMF is non-radiant. See Power Lines and Cancer FAQs at

http://www.mcw.edu/gcrc/cop/powerlines-cancer-FAQ/toc.html#4

 

[Warpspeed]

O/k, so let's say all the parameters for a complex power distribution system are entered into a 3D model, and some sort of plot of expected field strength versus frequency is generated.

Then I come along and plug in my heavy duty industrial arc welder into one of the branch feeds. What do you suppose the effect of that might be on the theoretical values throughout the whole system ?

It is difficult enough to test and measure radiation from a single closed "black box" entity. Something open, like a power distribution system it is just not possible to model because you do not know how clean the power coming in will be, or the nature of all the loads that may possibly be connected.

The power coming in is never clean, there are always switching transients, arcing insulators, remote lightning strikes, induced Rf voltages, all sorts of horrors not possible to predict for computer modeling.

All you can do is try to measure it after installation is complete, and hope.

 

[wrong100]

steven7
EMI stands for electromotive interference.
There are 4 basic types defined
1) conductive "on a wire"
2) conductive susceptable "subject to on a wire"
3) radiated? "energy in the air"
4) radiated sus "subject to energy in the air."

Each indepent item has its specs and the summation of these
must meet the total requirments.
The only items that should apply are 1 and 2.
and are controled by specs.
For someone doing power systems you have no control
over the line (which is what you are implying).
The only choice you have is to do resonant filter
or some such thing.

Boy did I get these out of order.
In short I think you are coping with only
P F (harmonic content), as the R F specs
address individule units.

For going to the power grid most energy fc above 10k
can be delt with a .1uf cap delta +line to nut configuration.
Fcc 15-49 defines the present requirements for com use.
Hope this helped.

 

[itsmoked]

EMI = ElectroMagneticInterference

 

[wrong100]

itsmoked
Sorry in strict terms you are correct.
The military lumps all types together.
Than seperates them as conducted and radation.
Fcc is the same?

The point is for power (AC IN) by the time you
get to the input mains, it "should" be just conduction,
which makes it a form of voltage, (current) either by
conduction or fields (as described by cuky2000).

Good example cuky2000

THink of this, (for testing) the mil started with one type "LISN" then to a 10u cap than to a different "LISN".
And each time the measurement concept changed.
This is only the major changes, it seems "when you understand and generate programs (computer) they
change the rules".
The last battle I lost on EMI was to a client who had speced the wrong LISN "line stabalization network".