Grounding VS Bonding

[__dan]

Teach me, Mike: how does the earthing connection limit the voltage to ground on equipment, bearing in mind that the definition of "ground" in the NEC is given as, "The Earth."?

Actually, I would say the earth ground limits the noise level on the grounding system. The equipment grounds are usually noisy at the source in big installations, with circulating ground loop currents, sometimes large ones. The grounding noise level is higher than the signal level of IT and data com equipment. The noise voltage is in the hundreds of millivolts range or less, but circulating ground loop currents can be tens to hundreds of amps, especially with large UPS's or dual fed switchgear.

The equipment ground requires you to bond back to the source, even if the grounding at that point is noisy. The earth ground requires you to connect to the earth, which is assumed or implied to be a zero volt, zero noise, reference point. At the service and at the origin of the SDS, there are these two different grounding conductors.

In a fault the equipment ground gets even noisier, but along the path back to the source carrying the fault current. The earth ground conductor is not carrying current in a ground fault, and so is clean of the noise on the equipment ground.

The earth ground point can be selected to perform this function, a clean and noise free reference point. It can also be missed that this function is necessary or important.

 

[dereckbc]

Actually, I would say the earth ground limits the noise level on the grounding system. The equipment grounds are usually noisy at the source in big installations, with circulating ground loop currents, sometimes large ones. The grounding noise level is higher than the signal level of IT and data com equipment.

Please enlighten me. I design and build grounding systems in extremely large data centers and telephone offices.

 

[George Stolz]

Start your own IT thread guys. :lol:

 

[don_resqcapt19]

I agree, even in a NEC compliant EGC is employed will not save your life. Even on a simple 120 VAC circuit, if there is a ground fault at the equipment, the chassis will have 60 volts until the fault is cleared. Now here is the shocker and killer. If that was say 480, you get hit with 240 or more volts during a fault.

Dereck, That sounds high to me. The rise in voltage between the metal parts of the faulted equipment and the earth should be equal to the voltage drop on the fault clearing path.

 

[Sahib]

Yep as I can see in this thread, grounding has a few folks confused. I have taught several grounding classes even wrote a book or two on the subject, and I make my living at it in the telecom sector. Trust me there are no bigger jerks than there is in the Telecom when it comes to grounding. Most of the equipment manufactures are clueless, especially the ones that claim they must have 5 ohms or less to earth impedance. What follows if 3 very simple circuit diagrams I have used that gets the basic idea across. Excuse the mess, but I drew it up real quick for here. Two circuits will work and safe, one will get you killed, and only one is compliant Can you tell which one and why they would or would not work?

The first one is dangerous to life, because no EGC. The third one is safe, as EGC present , The second one is code complaint as one source conductor is grounded Correct?
But in real life condition the safety depends the impedance ratio between phase and neutral-EGC combination. Higher is the ratio i.e making the neutral-EGC equivalent cross-sectional area bigger, lower will be the shock voltage. This is made possible by the Code by requiring to connect all non current carrying metal parts to the EGC making the equivalent cross-sectional area as large as possible.

 

[jwelectric]

Teach me, Mike: how does the earthing connection limit the voltage to ground on equipment, bearing in mind that the definition of "ground" in the NEC is given as, "The Earth."?

Ever take a voltage reading? That voltage is in reference to what? EARTH

You are hanging your hat on that section, stating it is more important than the bonding principle. You are adamant on this point.

Whoa up here just a little. You won?t be able to make a quote where I have said that one is more important than the other.

All of us have seen countless examples that this section in the NEC is patently unenforceable and defies the laws of physics, yet you seem to believe it more important than sections the majority fully understand and believe to be far more important at the equipment served.

No physics at all involved here but simple code verbiage that can be found in 250.4

So please, without posting miles of code we already have access to, enlighten us. Why is it more appropriately called EGC instead of EBC.

Because it connects all non-current carrying metal that encloses electrical wires and equipment to earth.
Bonding is something like ensuring electrical continuity around expansion joints in metal conduit, the metal boxes used in non-metallic cables, the jumper installed from a metal box to the metal yoke of devices where metal raceways are used.

Bonding rules are a lot more lax than equipment grounding so to change the name of equipment grounding to equipment bonding is going to change the rules of the earth connection and the potential of earth and a metal non-current carrying that is elevated above earth. Read 250.4(A)(1) and (2) for more information on the subject. When you read this for a few minutes forget about bonding and let nothing but earth be on your mind. Once this has sunk in real good then proceed with the bonding of your choice.

 

[roger]

Here's the way I see it, you have a "Grounding Conductor" that should be called an "Earthing Conductor" and you have an "Equipment Bonding Conductor" that is incorrectly called an Equipment Grounding Conductor.

You can lift the Grounding (should be Earthing) Conductor from the electrode and the "Equipment Grounding (should be Bonding) Conductor" would still be bonding everything together yet it would not be Grounding anything so, EBC makes more sense to me.

Roger

 

[Sahib]

The original intention was to keep an equipment enclosure at ground potential by connecting a conductor between the equipment and the ground. Thus the name EGC derives. But as shown in my earlier post, the EGC should be bonded with other non current carrying metallic parts to make its protective function more effective. So the conductors which bond the EGC should be called bonding conductors and not the EGC itself should be so called because it, otherwise, defeats its original purpose.

 

[ggunn]

What about a fourth case which is a combination of #1 and #2, i.e., where there is both an electrode at the load and an EGC back to the electrode at the source? Would that be more or less safe? Compliant?

Yes and Yes as stipulated by 250.96 However if it is sensitive electronic equipment may or may not interfere with proper operation of the said equipment.

EDIT NOTE

Sorry I goofed while trying to reply with quote. Please accept my apology. Being a MOD I can edit your stuff and did so by accident.

 

[jwelectric]

Ever take a voltage reading? That voltage is in reference to what? EARTH

Whoa up here just a little. You won?t be able to make a quote where I have said that one is more important than the other.

No physics at all involved here but simple code verbiage that can be found in 250.4

Because it connects all non-current carrying metal that encloses electrical wires and equipment to earth.
Bonding is something like ensuring electrical continuity around expansion joints in metal conduit, the metal boxes used in non-metallic cables, the jumper installed from a metal box to the metal yoke of devices where metal raceways are used. And by the way this is referred to as the equipment bonding jumper.

Bonding rules are a lot more lax than equipment grounding so to change the name of equipment grounding to equipment bonding is going to change the rules of the earth connection and the potential of earth and a metal non-current carrying that is elevated above earth. Read 250.4(A)(1) and (2) for more information on the subject. When you read this for a few minutes forget about bonding and let nothing but earth be on your mind. Once this has sunk in real good then proceed with the bonding of your choice.

. .

 

[dereckbc]

Dereck, That sounds high to me. The rise in voltage between the metal parts of the faulted equipment and the earth should be equal to the voltage drop on the fault clearing path.

OK Don assuming the EGC conductor is the same gauge wire, or even if smaller than the supply conductors we have a simple voltage divider circuit right?

If that is a true statement then we have equal lengths of the hot conductor and EGC, thus the voltage is roughly half the supply voltage. Now the kicker is do we have multiple parallel fault return paths (like conduit) which would lower the impedance, thus the voltage. But what about a home appliance like a dryer or fridge where all you have is the EGC.

My point is even a compliant NEC EGC there is no guarantee of life safety. Should protect life more than 99% of the time, but there is that small chance a person could be in contact at the moment of the fault. On a 120 VAC circuit, most likely make you jump and mess up your underwear. On a 480 well....

 

[dereckbc]

The first one is dangerous to life, because no EGC.

Correct, and is what I have to point out to equipment manufactures a 5 Ohm ground is meaningless and has nothing to do with clearing a fault. In this case you have 10 Ohms total and at 120 volts is 12 amps on a 20 amp circuit is a load, not a fault.

Earth resistance is pretty much meaningless with operating voltages of 600 volts or less. NEC forbids earth to be used as a conductor. Now when you move up to utility voltages earth impedance is a factor and can be used as a conductor and fault path. Utilities use earth as a fault path conductor with MGN systems. They even use earth as a return conductor on SWER systems.

The third one is safe, as EGC present

Mostly correct, but is the demonstration ground (earth) has nothing to do with clearing a fault. This would be an Equipment Bonding Conductor, not an EGC. In fact you can make a very good argument this is safer than an NEC earthed ground referenced system. Pretty close to what you have with a ship. automotive, and aerospace.

The second one is code complaint as one source conductor is grounded Correct?

Correct.

The whole point is earth is not required, just a planned fault return path capable of passing enough fault current of at least 6 times the OCPD rating. Once you wrap you brain around that, then everything falls into place and makes sense and you can connect the dots.

 

[__dan]

Please enlighten me. I design and build grounding systems in extremely large data centers and telephone offices.

We can talk about the source end or the utilization end.

At the utilization end, pallets of rack mounted equipment comes in and the manufacturer's instructions say "earth ground required". Commonly, there is a threaded stud at the back of the equipment enclosure provided for attachment, and recall, internally the equipment has either an SDS or electrical monitoring devices and signal level comm referenced internally to "ground".

I have an easy method compliant with 250.30 (A) 4, but it involves running an additional wire, which is I believe the intent. The earth ground system is already present in the raised floor, by tapping the GEC provided for the raised floor area step down distribution transformers. It is also at the common grounding and bonding busbar, but the busbar is less accessible.

What I see happening is the equipment "earth ground required" is largely ignored, including egregiously when the supplied equipment grounds have a measurable ground loop current on them violating 250.6.

If you see rack mounted equipment come in that says "earth ground required". Do you intend to run a wire (a GEC) in addition to the power supplied equipment ground, or is the equipment ground satisfactory by itself?

 

[__dan]

Here's the way I see it, you have a "Grounding Conductor" that should be called an "Earthing Conductor" and you have an "Equipment Bonding Conductor" that is incorrectly called an Equipment Grounding Conductor.

You can lift the Grounding (should be Earthing) Conductor from the electrode and the "Equipment Grounding (should be Bonding) Conductor" would still be bonding everything together yet it would not be Grounding anything so, EBC makes more sense to me.

Roger

I still don't get what this thread is about.

The only case where equipment or structures are not grounded is unpowered, no electric, systems. Other than small battery powered tools, if it has electricity, a "grounded" bonding system is required. Trying to separate the grounding from the bonding system fails what is required or specified. It is like buying a dozen eggs and complaining about the shells, you are buying or building a working system, the bonding is required to be grounded. If it is not grounded, there is some failure somewhere to meet specification.

Bonding to me is not adequately descriptive. A job that is not bonded, bonding is default insurance. Say "grounding" and people are now familiar with the usage, they think, "oh that's simple, any caveman can ground it". But not always as the thread shows, grounding is a specific circuit path, in addition to the mechanics of construction. Bonding is a word that is ripe for abuse, possibly implying the ground fault path, but not the earth ground path, when the "grounded system" requires both. Why would you try to imply the bonded system is not grounded.

 

[don_resqcapt19]

OK Don assuming the EGC conductor is the same gauge wire, or even if smaller than the supply conductors we have a simple voltage divider circuit right?

Thanks ...I was not looking at it that way.

...
My point is even a compliant NEC EGC there is no guarantee of life safety. Should protect life more than 99% of the time, but there is that small chance a person could be in contact at the moment of the fault. On a 120 VAC circuit, most likely make you jump and mess up your underwear. On a 480 well....

Yes, I agree that the EGC does not guarantee life safety and a connection to earth with out an connection to the system for a fault clearing path provides almost no protection, unless you are standing on the grounding electrode.

 

[don_resqcapt19]

Ever take a voltage reading? That voltage is in reference to what? EARTH

It is a very rare case when I take a voltage reading that I care about the voltage to earth. I am looking to see the voltage between the conductors of the system

...
Bonding rules are a lot more lax than equipment grounding so to change the name of equipment grounding to equipment bonding is going to change the rules of the earth connection and the potential of earth and a metal non-current carrying that is elevated above earth. Read 250.4(A)(1) and (2) for more information on the subject. When you read this for a few minutes forget about bonding and let nothing but earth be on your mind. Once this has sunk in real good then proceed with the bonding of your choice.

It will always remain my opinion, that the main function of the EGC is to bond the non-current carrying parts of the equipment back to the grounded conductor of the system, via the main bonding jumper.

The connection to earth does nothing to keep the connected equipment at earth potential. The only place that actually occurs is if you are touching the equipment that is connected to the EGC and standing on or near the grounding electrode.

 

[don_resqcapt19]

...
Bonding to me is not adequately descriptive. A job that is not bonded, bonding is default insurance. Say "grounding" and people are now familiar with the usage, they think, "oh that's simple, any caveman can ground it". But not always as the thread shows, grounding is a specific circuit path, in addition to the mechanics of construction. Bonding is a word that is ripe for abuse, possibly implying the ground fault path, but not the earth ground path, when the "grounded system" requires both. Why would you try to imply the bonded system is not grounded.

And I see this completely the other way...people see the word "grounding" and do think that any caveman can ground it...and they do, by making a connection to the earth without making the required "bonding" connection to the system.

The "bonding" connection that the conductor we now call the "equipment grounding conductor" provides is far more important in the safe operation of the system than is the connection to the earth that the EGC also provides. The use of the word "grounding" in the term "equipment grounding conductor" implies that the connection to earth is the most important function of that conductor, when in fact that part has very little importance.

 

[dereckbc]

We can talk about the source end or the utilization end.

At the utilization end, pallets of rack mounted equipment comes in and the manufacturer's instructions say "earth ground required". Commonly, there is a threaded stud at the back of the equipment enclosure provided for attachment, and recall, internally the equipment has either an SDS or electrical monitoring devices and signal level comm referenced internally to "ground".

I have an easy method compliant with 250.30 (A) 4, but it involves running an additional wire, which is I believe the intent. The earth ground system is already present in the raised floor, by tapping the GEC provided for the raised floor area step down distribution transformers. It is also at the common grounding and bonding busbar, but the busbar is less accessible.

What I see happening is the equipment "earth ground required" is largely ignored, including egregiously when the supplied equipment grounds have a measurable ground loop current on them violating 250.6.

If you see rack mounted equipment come in that says "earth ground required". Do you intend to run a wire (a GEC) in addition to the power supplied equipment ground, or is the equipment ground satisfactory by itself?

Everything you think you know has been tossed out the window with MESH GROUNDING. That is where all the equipment is installed on a Equopotential Ground Plane (installed on a Diamond Steel Plate) and all ground and return conductors are bonded together multiple times. In equipment racks the battery return is a short piece of wire connected directly to the equipment chassis. AC powered equipment is supplied with Point of Use inverter installed in the rack and b0th Equipment and Signal ground are the equipment track they are mounted in.

Today's communications is not a unbalanced topology like you rfather used with RS-232 or coax. Those days are long gone. Today all balanced topology that has no use for earth.

 

[__dan]

Everything you think you know has been tossed out the window with MESH GROUNDING. That is where all the equipment is installed on a Equopotential Ground Plane (installed on a Diamond Steel Plate) and all ground and return conductors are bonded together multiple times. In equipment racks the battery return is a short piece of wire connected directly to the equipment chassis. AC powered equipment is supplied with Point of Use inverter installed in the rack and b0th Equipment and Signal ground are the equipment track they are mounted in.

Today's communications is not a unbalanced topology like you rfather used with RS-232 or coax. Those days are long gone. Today all balanced topology that has no use for earth.

Well, in my father's data center, there is a lot of legacy coax and serial cable, in addition to cat 5.

The typical data center has the raised floor grid, and the steel structure of the floor would be that mesh you refer to, except that the floor grid is not UL listed for grounding per 250.118, so you can bond to the floor, but not ground to it.

I do see what you are saying about connecting to a site built equipotential plane. What I am saying is in a legacy data center that equipotential plane is the grounding electrode system, preferably tapped at a point remote from the noisy sources like the UPS and the dual fed main switchgear. Reference your ground to a point that is electrically quiet.

I have seen the ground loop currents strong enough to cause piles of board and port failures in modern blade, cat 5 connected equipment. It was in a room I cited for violating by omission of 250.30 (A) 3, the PDU's had equipment grounds but no GEC's.

And if you wish to, I would like to know if you require a GEC in addition to the equipment ground, when the manufacturer's instructions say "earth ground required (and there is no mesh ground structure).

 

[dereckbc]

And if you wish to, I would like to know if you require a GEC in addition to the equipment ground, when the manufacturer's instructions say "earth ground required (and there is no mesh ground structure).

Start a thread and I will discuss. But let's not hijack this thread.