Trying to understand EGC & local ground

Status
Not open for further replies.
he is saying its not intended for clearing faults but it's for personel protection.

it wouldn't do much in terms of that either would it?
 
081209-2008 EST

The NEC statements quoted by Pierre I believe mean the following:
(1) You must have a dedicated conductor(s), called an ECG, running from every piece of equipment back to a central common bus at the service entrance.
(2) You can not use the building steel structure in place of said dedicated conductor.

This does not say that the building structure can not be tied to this same common bus at the service entrance. It does not say that the building structure can not carry fault current. It does not say how much fault current is allowed to flow thru the building structure.

It is quite possible that if the building I beams are welded together that this will provide a much lower resistance path for fault currents than the required EGC. It will also provide much lower resistance between machines and thus likely less noise voltage between machines than would the required EGC. But this does not mean it can be used as the required EGC.

Building steel may not be a reliable ECG path in comparison with a known continuous length of copper wire of an adequate size, and generally in a protected enclosed path. This is because building steel is intended for structural purposes, is made of many small pieces and corresponding joints, and there are no specifications on how to build this structure as a guaranteed low impedance path. It might be good and it might not.

.
 
gar said:
081209-2008 EST

The NEC statements quoted by Pierre I believe mean the following:
(1) You must have a dedicated conductor(s), called an ECG, running from every piece of equipment back to a central common bus at the service entrance.
(2) You can not use the building steel structure in place of said dedicated conductor.

This does not say that the building structure can not be tied to this same common bus at the service entrance. It does not say that the building structure can not carry fault current. It does not say how much fault current is allowed to flow thru the building structure.

It is quite possible that if the building I beams are welded together that this will provide a much lower resistance path for fault currents than the required EGC. It will also provide much lower resistance between machines and thus likely less noise voltage between machines than would the required EGC. But this does not mean it can be used as the required EGC.

Building steel may not be a reliable ECG path in comparison with a known continuous length of copper wire of an adequate size, and generally in a protected enclosed path. This is because building steel is intended for structural purposes, is made of many small pieces and corresponding joints, and there are no specifications on how to build this structure as a guaranteed low impedance path. It might be good and it might not.

.
Click to expand...

I agree that the building steel could have less of a resistance path than an EGC but only could act as a EGC efficiently if it was bonded all the way back to the source, since we've discussed that due to ground resistance it most likely wouldn't clear faults with current flowing to ground.

In addition to your comment regarding less noise voltage between equipment due to less resistance, I would also think that it would lower the touch potentials between equipment for the same reason. By bonding equipment to grating or steel floor it would create a low resistance equalpotential plane and reduce touch potentials between equipment.
 
The code section posted does not permit the structural steel to be used as an EGC.


Think of this scenario and you will understand why this may be so.


The electrician uses the steel as his/her EGC. There is a fault condition, but the fault condition is such that a limited amount of fault current travels along the steel, hence the time for the OCPD to open takes 5 minutes. During that time, the 277v circuit fault current is imposed on the structural steel of the building....does that sound like a good idea?
The possibility of having fault current on structural steel is not such a great idea, regardless of how long it is flowing in the steel.
 
081209-2342 EST

mull982:

The resistance of building steel to the earth would have nothing to do with its effectiveness as a fault path. If the building steel is bonded to the neutral bus at the sevice entrance, then it is the resistance of this path that determines how much current flows thru the steel during a fault.

The obvious reason that this is not allowed as the EGC is because there is no control on how good are the many joints in the structure. But in many cases its resistance back to the main panel may be lower than the actaul EGC. It all depends upon the structure, and aging of the joints.

On BB64 where I served for a while the resistance of the deck just below my bunk was probably pretty low. It was rather thick armor steel, maybe 8" to 12" thick.

.
 
081209-2355 EST

Pierre:

Unless you insulate all equipment and connecting elements there will be fault currents that flow thru structural steel.

In an automotive plant you will have various things like electrical ducts, water lines, air lines, lube lines, fuel lines all loosely bonded to the structural steel and the machines. Machines are in many cases mounted on steel pads in the concrete floor. All of these interlinked things form one big conductor back to the service entrance. Probably at least thru the electrical conduit near the main panel all this low resistance structure connects to the neutral and ground buses at the service entrance.

That does not make it a reliable EGC. But it will conduct some of any fault current.

.
 
gar said:
081209-2342 EST

mull982:

The resistance of building steel to the earth would have nothing to do with its effectiveness as a fault path. If the building steel is bonded to the neutral bus at the sevice entrance, then it is the resistance of this path that determines how much current flows thru the steel during a fault.

The obvious reason that this is not allowed as the EGC is because there is no control on how good are the many joints in the structure. But in many cases its resistance back to the main panel may be lower than the actaul EGC. It all depends upon the structure, and aging of the joints.

On BB64 where I served for a while the resistance of the deck just below my bunk was probably pretty low. It was rather thick armor steel, maybe 8" to 12" thick.

.
Click to expand...
Actually it is the impedance of the path and that is an issue with remote paths in an AC system. On high current faults, above 200 or 300 amps, the inductive reactance will limit the current that flows on a remote ground fault path such as the building steel. In extreme cases with only a "remote" fault clearing path, the inductive reactance may limit the fault current enough that the OCPD does not quickly act to open the circuit.
 
O.K. So I see understand and see the reasons why a structure cannot be used as an ECG.

My lingereing question is, then why connect the building steel to ground rods. In the origonal post with the light pole and the supporting documents the followed it showed that this served no real purpose.

So why bother driving a ground rod and attaching them to every so many numver of cloumns?
 
High Imp. Grounding?

High Imp. Grounding?

No one has asked the question if the OP's facility uses a High Imp. grounding system. If that is the case the facility may not "trip" immediately on a GF...in fact it may run for hrs before the process is shut off & the fault is located. In many large industrial facilities the structural steel is used as an "alternate" path for grounding in case the EGC to a large motor fails (connection fault). The structure is connected to ground to in an attempt to create a more equipotential ground plane - yes, in a effort to reduce shock hazards - the example above of a ground rod griven in the ground for a light pole is NOT the same situation as a large industrial facility that has its structual steel connected to ground. In fact, MSHA CFR 30 Part 77 requires that the frames or enclosures of all electrical equipment "must be connected to ground in a method that insures that there is no potential between the earth (interpreted as local area) and said metal enclosures" - In 98% of my facilities 480 vac systems are resistance grounded to 1 ampere or less - we will run with a GF for up t0 24 hrs - it is imperative that we maintain as low a "step potential" as possible to reduce potential shock hazard. We use 3 methods of ground - EGC, Structural Steel, and conduit - any two of these must be in working order at anytime - yes, the EGC is the "primary" ground fault path, but these are 24 hr /7 day facilities so if the EGC connection or wire fails- we still have a "reference" to ground. we must have an external Jumper from Steel to motor or enclosure or we recieve an instant violation (based on interpretation). There are many IEEE publications on these apllications as well as a book "Safety Grounded Systems For Mining" by Gary Powell that explain these in detail. In a resistance grounded system the resistor holds the largest voltage drop (simple resistor divider circuit) For example if a 480 system has a 277 ohm resistor between transformer neutral and ground and the EGC has 1 ohm of resistance, a GF at this location would create a current of 0.996 amps - or a voltage difference between "Earth" and this device of 0.996 volts (276 volts across the neutral resistor) - this is assuming the full voltage difference (ohmic) between the device and ground (shock potential) now lets say we have structual steel and conduit grounded as well & a jumper to the motor - now the shock hazzard is the GF currrent times local resistance to local ground (Earth) - on a GF if the local ground grid connection to the faulted device is .5 ohms then the local shock hazzard is reduced to less than O.5 volts. If the Local resistance to ground is 0.1 ohms (typical) then the shock hazzard is less than a tenth of a volt.
 
081210-1108 EST

don_resqcapt19:

It is the impedance that is the determining factor. Is the impedance less for the loop of the EGC and its adjacent hot wire that was shorted to the EGC, or for the loop of the hot wire and the large distributed conductive structure of the building. I do not know. It depends on when inductive reactance becomes dominate over resistance.

mull982:

The A-frame building you mentioned probably had steel beams forming the A and these are thus on the outside or close to the outside of the roof. An easy target for lightning. If these beams are insulated from earth, then it sort of makes sense to provide a ground rod or grounding system for each A assembly. Anything that you can do to get the lightning to stay outside the building is a benefit.

In reviewing some of the references I provide on my web site I found that the one I wanted to present here is a broken link. It had to do with using broad wide conductors from the grounding electrodes to what is to be grounded for lightning protection.

There is another reference that still exists on grounding CNC machines which I shall include here.
http://ecmweb.com/mag/electric_cnc_machine_tool/

.
 
Status
Not open for further replies.
Top