Low resistance to earth, is it necessary?

Status
Not open for further replies.
corvalan

corvalan

Member
The reason for this email is to have your opinion regarding the importance of having low impedance to earth.

Very often, I found in the industry the requirement that a grounding electrode system should have very low resistance to earth, sometimes even approaching the magical one ohm.

I find this unreasonable under a practical and also under a theoretical point of view.

This subject is complex to analyze in its entirety, but I believe that if we restrict the subject to only grounding (without considering lightning strikes and without considering sensitive electronic equipment), we can conclude the following:

If the building has a well designed and implemented grounding system according to the NFPA publications and IEEE standards, there is no need for the grounding electrode system to have a low resistance.

If we now include lightning strikes, I believe that the subject changes. I agree that low impedance to the lightning current will help prevent side flashes from the down conductors to grounded metal equipment near it. But the NFPA 780-2000 considers that it is much more important to prevent side flashes by having frequent and excellent bonding instead of low impedance to ground. So much is this the case, that the NFPA DOES NOT REQUIRE minimum impedance neither a minimum resistance to earth for their grounding terminals. So again, you can have an excellent grounding system with grounding electrodes having high resistance to earth.

Bringing electronic sensitive equipment to the table, is more complicated, and will require possible more time that you have assigned to answer these questions. But I believe that, even when the building houses sensitive equipment, if you have a well designed grounding system and a well designed lightning protection system, the resistance or impedance to earth of the grounding electrode system is irrelevant.

Please engage in this discussion, because I believe that there is too much mystery waiting to be solved in this subject.

Ruben Corvalan
Director
CYAMSA
 
Re: Low resistance to earth, is it necessary?

Maybe my knowledge is too limited but isn't part of the reason for the low impedence to protect persons? Since a person may be grounded at the site by virtue of the fact that he is actually standing on the ground, without a low impedence connection any ground fault potential to earth might choose him as a conductor. I have heard, although never actually experienced or proved it, that two points of earth contact can have some rather dramatic differences of potential and by using the grounding electrode system, these potentials are reduced.

Am I out in left field?

Bob
 
Re: Low resistance to earth, is it necessary?

Ruban, I am pressed for time at the moment and cannot take a lot of time to engage in a conversation. But for now, a lot of it has to do with not letting the contractor/electrician only perform minimal NEC requirements of driving two-rods, connect then with 6 AWG using mechanical connectors and calling it a day. It is inadequate to use as lightning protection and prone to failure.

Chew on this for now. What is the impedance of a 10-foot long piece of 750 MCM and 6 AWG @ 10 MHz, connected in series with a 5-ohm electrode @ 100 Hz. More later.
 
Re: Low resistance to earth, is it necessary?

Bob:

If lightning hits the building and the person is standing outside the building (touching earth with his feet), REAL close to the grounding terminal of a lightning protection subsystem, he will be subject to a large potential gradient across his feet. This potential gradient his called step potential and could be several thousands volts but extremely short lived, may be 50 microseconds. I doubt that engineers specifying low earth impedance are doing it to avoid this situation.

I did not understand the sentence: "without a low impedance connection any ground fault potential to earth might choose him as a conductor". The comment I can give you regarding this is that it is convenient for the person standing outside to have the highest possible impedance to ground. This will minimize the magnitude of currents passing through his body.

With respect to your final sentence, I have this comment. These steep potential gradients can me smoothed by installing a ground grid below the ground, like those in external substations. The regular grounding electrode system of a building is not designed nor does a good job in preventing these steep gradients in the grounds outside the building during lightning strikes.

Thanks
Ruben
 
Re: Low resistance to earth, is it necessary?

Ruban, I am not sure what exactly you want to discuss. There is safety (Power frequencies) grounding which is the realm of the NEC, and performance grounding (high frequency, noise, etc) which is the realm of equipment engineers. When I design protective grounding systems for telephone, radio and data centers I break it down into sub-systems. Each sub-system has its particular function. The sub-systems are:
? Grounding electrode system.
? AC Equipment ground
? Lightning protection / Discharge ground.
? DC equipment ground.
? Signal grounds.
? Isolated ground planes.
? Integrated ground planes.

NEC minimum requirements mostly deal with the first two sub-systems and to a lesser degree the third sub-system. While performance engineering includes NEC requirements but go well beyond.

Since you mentioned lightning I will start there. For most residential, commercial, and some industrial applications a low impedance ground is not necessary. These types of installations typically employ a two-rod and water pipe electrode system. They do not offer much of anything in terms of lightning protection. Once you consider the impedance of the GEC and EGC at lightning frequencies, the GE impedance is irrelevant as you stated. Going back to my earlier post what is the impedance of a 10-foot piece 750 or 6 AWG (GEC) @ 1 MHz and 10 MHz connected to a GE of 5 ohms. Well you get 260 and 2.6 K ohms. So it does not really make any difference if the GE is 5 or 25 ohms, you are limited by the GEC. Now add EGC?s to the equation and your impedance to earth goes off the chart at high frequencies

With a two-rod and water pipe system, even if it were say 5 ohms still would not be able to dissipate lightning quickly because of ground saturation, causing step gradience. However this can be over come by use of ground rings, grids, and radials to effectively form multiple loops to lesson the inductive impedance and shorting out the step potential difference.

That is enough for now, where to next?
 
Re: Low resistance to earth, is it necessary?

Dereck:
The purpose for this string is to find out if it is necessary a low impedance to earth for certain applications.

I have the belief, that in any application in an industrial environment, if you have ADEQUATE bonding according to the NFPA codes and IEEE standards, the impedance to earth is irrelevant.

So lets still remain in the industrial environment with a building with proper lightning protection and proper grounding.

Lets say that a lightning bolt of 10,000 A strikes the building. The energy of this bolt of lightning will be coupled to the building in three different ways. Ohmic, Inductive and Capacitive.

Lets start with Ohmic. Assume that the lightning current finds a very high impedance to earth of 100 ohms. This creates steep gradients on the local earth outside the building. This creates a potential difference of 1,000,000V from all metallic equipment inside the building to remote ground.

I belive this situation will not create any problem because every metallic equipment inside the premises has raised to the same voltage (1,000,000V)and therefore do not have voltage differences between them.

The steep gradients will be on the earth on the periphery of the building and not in the inside. So why compensate for this by providing lower impedance?

Now with respect to the inductive coupling. If the there is any conductive loop near the down conductor, there will be inductive coupling. I belive that the best way to compensate for the overvoltages this will create is installing SPD's in all energy and data ports of electrical and electronic equipment. Not decreasing the impedance of the lightning ground terminals and installing the sensitive electronic equipment far away from down conductors.

With regards to capacitive coupling, which I belive is the weekest coupling, similar me assures should be implemented as the previous paragraph.

So good bonding will prevent damage from ohmic coupling and good overvoltage protection design and implementation will prevent damage due to inductive and capacitive coupling. Therefore, is my impression, that lower impedance to earth is irrelevant.

Please continue with this discussion.

Thanks
Ruben Corvalan
 
Re: Low resistance to earth, is it necessary?

Originally posted by corvalan:
Dereck:
The purpose for this string is to find out if it is necessary a low impedance to earth for certain applications.
Click to expand...
IMHO yes. Particularly radio and telephone applications.

Raw lightning sources have a high source of impedance, about 5000 ohms. Therefore, when lightning strikes, the current delivered by the stroke will "force" its way through the building to earth, developing voltages that are proportional to the impedance through which the current flows with little or no change in current magnitude. Thus lowering the impedance of the building results in lower voltages throughout the building, reducing personnel hazards and the likelihood of damage or misoperation of equipment.

There are techniques to eliminate or limit the voltage potential differences created. One popular method in telecom/radio is by using a "ground window" forming a "single point ground" and incorporating an "isolated ground plane". By using this technique no voltage difference could possible be generated across this point during short circuit occurrences in an ac or dc system when lightning current flows in a building.

Overall I agree with your assessments providing certain techniques like a "ground window" with "isolated ground planes" are used to limit voltage stress during lightning events.

For the most part, all installations use an "integrated ground plane" approach. This is the system you are speaking of by proper bonding, forming multiple loops, and multiple connections to earth via a grid or ground ring. In this application a typical two-rod and water pipe ground electrode would be practically useless IMO. So to make it effective, a ring or grid would be needed. If you are using a ring or grid ground electrode system, it would stand to reason the impedance would be much lower than a rod and pipe system. Not too mention the added protection a ring or grid offers from the equipotential ground plane and minimizing step potential differences.

I have searched long and hard where the origin of a 5-ohm ground spec come from. The best I can tell it either came from the telephone/radio industry, or from military bomb storage installations via Herbert Ufer. The telephone industry once used earth as a signal path for ring voltage and trunk start indication, so a low impedance to earth was necessary at that time. Since then earth is not used, and the spec has remained in place, but still has value in the industry to provide protection in cable entrance facilities, and radio tower grounding.

Edited for clarification and spelling errors.

[ June 12, 2004, 11:31 AM: Message edited by: dereckbc ]
 
Re: Low resistance to earth, is it necessary?

Thanks for keeping this theme alive.

My answer to your comments, assume that there is a lightning protection system and a grounding system designed and implemented according to the NFPA codes and IEEE standards.

1. With respect to your comment:

Raw lightning sources have a high source of impedance, about 5000 ohms. Therefore, when lightning strikes, the current delivered by the stroke will "force" its way through the building to earth, developing voltages that are proportional to the impedance through which the current flows with little or no change in current magnitude.

I agree 100%

2. With respect to your comment:

Thus lowering the impedance of the building results in lower voltages throughout the building, reducing personnel hazards and the likelihood of damage or disoperation of equipment.

I have the following comments:

If lower voltages throughout the building mean lower voltage differences between metallic bodies throughout the building, I disagree with your comment. The differences of voltages between metallic bodies (panels, conduits, structures, conductors, shields, etc) do not change if the impedance to earth is lowered. And lowering the impedance of earth will not reduce the personnel hazards and the likelihood of damage of damage of equipment. What makes equipment to fail and what make a dangerous situation are voltage differences and as I said, lowering the impedance to earth does not affect these voltage differences in any way. Remember, I am assuming a well implemented grounding and bonding system according to NFPA and IEEE. I have been wrong before and I might be wrong now. If that is the case, please explain the process of why lowering the impedance to earth, the internal potential voltage differences between metallic components are increased.

I lower voltages throughout the building mean lower voltage of the metallic components with respect to remote earth, I agree with that comment. The higher the impedance to earth, the higher the voltage differences between the metallic bodies in a building and the remote earth. But I disagree that this higher voltage difference produces a higher personnel hazard for people inside the building and the likelihood of damage to electronic equipment installed inside the building. Remember, that no mater what is the voltage difference between the metallic components and remote earth, the voltage differences between the metallic components themselves are constant. Again, I would appreciate an explanation on how increasing the voltage of metallic components with respect to remote earth, will increase the hazards INSIDE a building.

I am aware, that if a person is standing outside the building, in the immediate vicinity of a ground terminal, higher impedance to earth will produce a higher step potential. Depending of several additional factors, these high gradients could be hazardous. This is the only advantage I see for lowering the impedance to earth. But I do not think that all the pressure for having low impedance to ground is being driven from this fact alone nor is the most important.

3. With respect to your comment:

There are techniques to eliminate or limit the voltage potential differences created. One popular method in telecom/radio is by using a "ground window" forming a "single point ground" and incorporating an "isolated ground plane". By using this technique no voltage difference could possible be generated across this point during short circuit occurrences in an ac or dc system when lightning current flows in a building.

I have the following comments:

I agree with you that there are several techniques to reduce the voltage differences between metallic components in a building. Ground window, isolated ground planes, single point grounding, etc. All of these methods include interconnections between different groups of metallic bodies. Some specific types of metallic bodies are tied together but not to other types of bodies, like forming interconnected subsystems. Eventually all these subsystems should be connected to the building ground. But again, the method for reducing the voltage differences here is not lowering the impedance to earth, but a specialized technique of bonding metallic components.

4. With respect to your comment:

Overall I agree with your assessments providing certain techniques like a "ground window" with "isolated ground planes" are used to limit voltage stress during lightning events.

I have the following comments:

I agree with this statement. But to keep it focused to our issue in hand, the benefit here is the specialized bonding techniques, not the lower impedance to earth. My statement is that the impedance to earth is irrelevant if you have a proper lightning protection system and a proper grounding system according to NFPA codes and IEEE Standards. I am not debating that specialized bonding techniques provide benefits.

5. With respect to your comment:

For the most part, all installations use an "integrated ground plane" approach. This is the system you are speaking of by proper bonding, forming multiple loops, and multiple connections to earth via a grid or ground ring. In this application a typical two-rod and water pipe ground electrode would be practically useless IMO. So to make it effective, a ring or grid would be needed. If you are using a ring or grid ground electrode system, it would stand to reason the impedance would be much lower than a rod and pipe system. Not too mention the added protection a ring or grid offers from the equipotent ground plane and minimizing step potential differences.

I have the following comments:

I agree with this statement. But to keep it focused to our issue in hand, the benefit (lowering the voltage differences between metallic bodies inside the building) is created by the bonding technique (ground ring), not the lower impedance to earth. The lower impedance to earth (and therefore lower step potential) that the ground ring creates is a secondary effect. In other words, you would obtain exactly the same benefits (lowering the voltage differences between metallic bodies inside the building) if you would install a ground ring, even if this ground ring does not create a lower impedance to earth.


6. With respect to your comment:

I have searched long and hard where the origin of a 5-ohm ground spec come from. The best I can tell it either came from the telephone/radio industry, or from military bomb storage installations via Herbert Ufer. The telephone industry once used earth as a signal path for ring voltage and trunk start indication, so a low impedance to earth was necessary at that time. Since then earth is not used, and the spec has remained in place, but still has value in the industry to provide protection in cable entrance facilities, and radio tower grounding.


I have the following comments:

I agree with you that the majority of the specifications for low impedance to earth, valid at a remote past) are pass-overs, if that expression exists. They are copied without really thinking. There are very specialized applications where low impedance to ground would be preferable. Especially in radio tower applications. But still in these applications, bonding and overvoltage protection and proper grounding (not earthing) are much, much more important.
 
Re: Low resistance to earth, is it necessary?

Ruban, I think we are on the same page. I will try to clarify. Anybody else, please jump in.

Ruban said:
If lower voltages throughout the building mean lower voltage differences between metallic bodies throughout the building, I disagree with your comment.

Comment:
If you have a properly installed lightning protection system, it basically forms a Faraday cage, which in turn lowers the overall ?building impedance? and shunts the fault current around the building structure, thereby reducing potential differences on metallic objects inside the structure.

I do not normally specify target earth impedance, rather design a system and take what I get. For the most part I get something 5-ohms or less from the design and installation specification. There are some sites like DOD, FAA, and Radio towers where I will specify a target impedance, but they are the exception rather than the rule. I agree with you that the earth impedance is irrelevant with a few exceptions already mentioned. The point I was trying to make is there are some types of installations (radio, communications, petroleum, and industrial process to name a few) that require a engineered electrode, lightning, TVSS, grounding, and bonding system and should not be left up to a contractors discretion using only minimum code guidelines.
 
Re: Low resistance to earth, is it necessary?

Dereck:

Thanks for your answer. I also think we are in the same page.

I still have a great interest in knowing why in those specialized applications (like radio towers) you spec low impedance to earth. What makes these installations different from the normal ones and why a low impedance to earth will benefit them?

Ruben Corvalan
 
Re: Low resistance to earth, is it necessary?

I'm just going to admit that I cannot help with the more theoretical aspects of this discussion since I am an electrician rather than a physicist. I have done a fair amount of grounding work and I have become a bit of a fanatic for lower impedance grounding based on having used lower impedance grounding to cure noise and repeating equipment failure problems on communications installations such as data networks and building telephone systems. In two cases bonding the failing system to the building ground did not clear up the problems but deep driven rods, metal well casing grounds, or ground rings did clear up the problem. To me this is counter intuitive. When the grounds of the telephone, computer network, and power systems are bonded together the problem of destructive surges between systems should have gone away. Why then did the problem go away in several places only after I used a well casing or deep driven rods as grounding electrodes.
--
Tom Horne
 
Re: Low resistance to earth, is it necessary?

Tom I would have to guess common mode noise. In telephone and data offices there are ground circuits used for signal purposes called either Signal Ground (SG) or Logic Ground (LG). Modern offices with switches use a SPG isolated LG which is reference to the battery positive bus or Xo of the transformer. The SPG and isolated nature make it impossible for common mode noise current to flow, and does not even have to be reference to earth, only referenced to the return of the power source. What I find is installation technicians will corrupt the LG by adding another earth ground, or connect it to an equipment frame.

The other method is by forming a grid of conductors on about 4-foot center, and multigrounding the grid in as many places as possible. This creates a very low high frequency impedance between connection points and too earth. The theory being this system is create such a low impedance that voltages developed by the common mode current will be below critical thresholds. So yes having a low earth impedance can help.

I cannot answer your question directly without knowing the particular's. Both methods of are used and each has it's advantages over the other.
 
Re: Low resistance to earth, is it necessary?

Dereck;

I believe that there must be confusion between low impedance to EARTH (accomplished by a very effective grounding electrode system) and low impedance to GROUND.

I believe that low impedance to EARTH DOES NOT MATTER at all to eliminate or mitigate common mode noise (that is, noise voltages that are common between signal conductors and GROUND or EARTH) in highly sensitive electronics circuits. But a low impedance to GROUND (example a low impedance in the high frequency grid installed in the area of sensitive electronic equipment) is EXTREMELY important to reduce the stray voltages that corrupt the communications between electronic equipment.

I do understand that there could be also a common mode noise voltage between the signal conductors of sensitive electronic equipment and EARTH. But due to the distance between the signal conductors and EARTH, and due to the fact that signal currents do not use the EARTH as a conductor, the interaction between them is negligible and irrelevant.

Not so between the signal conductors and GROUND. The grounded metallic components of a building and the signal conductors are very near and capacitive and inductive couplings are very important. That is why we need to have a very low impedance to GROUND (not to EARTH) so that stray voltages will produce high stray currents (due to very low GROUND loop impedances) therefore reducing the stray voltages (noise) between sensitive electronic equipment. Again, this low GROUND impedance (example a high frequency grid) is TOTALLY independent of the impedance to EARTH.

I believe (I might be wrong) that the EARTH does not provide the return path for any modern communication signal currents, therefore the impedance to EARTH is totally irrelevant.

I also believe that by using sophisticated BONDING techniques (not LOW IMPEDANCE TO EARTH techniques) you can minimize the ground loops that can be coupled into your signal circuits. Some of these sophisticated bonding methods require ONLY ONE connection to the BUILDING GROUND (structure, panels, conduits, etc.) or to EARTH. When you connect to EARTH or to BUILDING GROUND in more that one place, you may couple stray currents to your signal through loops that involve the EARTH or BUILDING GROUND as a conductor. But again, these are BONDING issues, not EARTHING issues.

Finally, with respect to the high frequency grid you mentioned with the 4-foot center. The grid ITSELF is designed to have a very low impedance to the circulation of a range of high frequency noise currents. This grid has to be connected to the BUILDING GROUND. But the impedance of this high frequency grid to EARTH is IRRELEVANT. The performance of this grid depends on the impedance of the grid itself to high frequency currents and not to the impedance of this grid to EARTH whatsoever. When the grid has low impedance to the high frequency noise currents, the common mode noise voltages between electronic equipment will be below the critical thresholds that you mentioned, but again this is totally independent of the low impedance to EARTH and low impedance to EARTH will not help in any way.

With respect to Tom`s comment. The reason of why the noise problem disappeared, not by an adequate bonding technique, but with the creation of a low impedance path to EARTH is unknown. It lacks theoretical support and goes against all my practical experience in the field. Therefore, I have the same response as Dereck`s, that is, I need to know the particular`s.
 
Re: Low resistance to earth, is it necessary?

Ruban, I agree with you 90%. The SRG is used to form a high frequency/low impedance path between pieces of equipment and the power source, and only needs to be bonded to the GES at one point for most installations.

However IMO where brute force RFI is involved, multiple connections of the SRG to the GES buried directly below the SRG using copper straps has been the only effective means I have witnesses in minimizing RF voltages induced from nearby transmitters. I am not saying it will eliminate the common mode noise, just minimize it enough below critical thresholds. Other than that I agree with you.
 
Re: Low resistance to earth, is it necessary?

Mike Holt in one of his news letters a while back, had an article about a bridge in NYC that workers (painters) was receiving shocks from. It was found that a near by AM radio station's antenna was a 4 element beam aimed right at the bridge. they tried all types of grounding. but the voltage was still there. I't amazes me that not many in the electrical industry understand that high frequency magnetically induced voltages can go right through a very low resistance load that would short out a DC or even a 60hz voltage. A typical base CB antenna has a shunt coil at the base of the antenna, that has a resistance of about .015 ohms but the ac voltage still reaches the top of it and radiates out. This can be explained via the skin effect that high frequency voltage will not only travel beyond the short but magnetically couple to any other metal that might be close enough to the resonate frequency of the voltage. This is one of the reason's lightning can induce voltage into a Faraday cage and still damage electronics within it. The only true way I can figure to stop the migration of this voltage since the base frequency changes each time lightning strikes, would be to design a very broad banded grounding system that would effectively be resonate to all the possible frequency's that lightning could impose. In some experiments we done in Gainesville, Fl. was multiple length ground radials from a single point Axis weather or not they were connected to earth didn't seem to matter as the lightning still struck them. starting at 18" in length up to about 180' and the longer the radial the higher off the ground we strung them. using ordinary surplus flash bulbs to detect the current path we noticed that only certain bulbs were flashed after a strike and each strike would trip a different set of bulbs.
We were using this experiment to try to understand why lightning would sometimes strike a tree that was only a few feet from an all metal tower that was hundreds of feet higher than the tree. the tree having branch's of deferent lengths (resonate) was the answer.

Nicoli Tesla was using a commutator spun at a high rpm on the output of a vandergraf generator to get an AC voltage at a very high level and frequency, when all of a sudden a lightning strike hit his equipment on a clear day in NYC. there after he was told to do his experiments elsewhere.

I just thought I'd throw some different thinking into this.
 
Status
Not open for further replies.
Top