This article was posted 11/26/2008 and is most likely outdated.

Grounding and Bonding for an LPS and SPDs
 

 

Topic - Grounding vs Bonding
Subject - Grounding and Bonding for an LPS and SPDs

November 26, 2008
This newsletter was sent to 29280 newsletter subscribers

Ask a Question |  Weekly Code GraphicQuizzes |  Free Stuff InstructorsOnline Training Products | Seminars | SubscribeUnsubscribe
[ image1 Post Comments | View Comments | Notify Me When Comments Are Added ] Web Page Version [Printer-Friendly]    

  

Grounding and Bonding for an LPS and SPDs

The following is an informative post by Ron Hotchkiss from Yahoo’s Lightning Protection Group:

 

Recently, there have been a number of discussions and inquiries regarding the grounding and bonding requirements and methods for use with lightning protection systems (LPS) and surge protective devices (SPD). Questions are often raised about the need for having a very low resistance ground for use with these systems.

The concept of having a very low ground resistance in conjunction with an LPS and SPDs is a topic that garners much debate amongst experts in LPS systems and SPDs. Of course, a number of suppliers, manufacturers, and installers promote a very low ground resistance as being paramount to the operation of such systems.

There is, of course, merit in having a very low ground resistance. Particularly in an LPS, low grounding impedance (notice the change from "resistance" to "impedance") may aid in the prevention of arcing due to a large difference in potential that may be developed because of a "higher" impedance connection to the earth. Poor connections or even sharp bends can present higher impedance at lightning frequencies. This scenario may be worsened if the distance between the source of the disturbance and the SPD is large or if the impedance is high. This impedance (and resultant voltage drop) prevents the SPD from being able to limit the potential to an adequate level. Of course, sharp bends are not desirable when constructing a lightning protection system. The inductive voltage drop across a sharp bend may exceed breakdown of the air between its de facto "terminals" – that is, the arc can "by-pass" the down conductors of an LPS either to another location on the down conductors or to the structure itself if the lightning current encounters a "high" impedance bend in the conductor or in a connection. Either case can create damage to the structure or fire.

These statements highlight the issue of connection lead length and sharp bends in both the lightning protection system (LPS) and in the connection of the SPD to the electrical system. A bend in the conductor may not seem significant, but at the frequencies involved it certainly is as impedance is directly proportional to frequency (consider the rise time associated with lightning wave fronts are often less than 10 microseconds – re-strikes are typically less).

Further regarding the grounding connection, it should be understood that protection can be achieved even without the presence of low impedance ground through proper bonding of the electrical system. This statement would seem to be contrary to what is often emphasized, but please keep reading. The most important concept, in my opinion, is voltage equalization as opposed to current diversion – although they go hand in hand. One must consider that there are locations on the Earth where a low impedance ground is not achievable – for example dry, arid locations like those in the southwest United States. In locations such as these, the "grounding" system of the electrical system is really more of a "bonding" system. The key to preventing damage to a structure, electrical system, or equipment connected to the electrical system is to create a situation where the grounding, bonding and protective systems prevent a difference of potential between points in the electrical system. Without a difference of potential, no current can flow as dictated by Ohm's law or in reverse, if current flows without the presence of significant impedance, a significant difference of potential will not be created. My statements are not intended to "de-emphasize" the concept of grounding – but to further "emphasize" the importance of good bonding.

With regard to the testing of grounding connections (i.e. the impedance of the ground rod(s) connected between the lightning protection system and earth), many devices used to make these measurements may not be making these measurements at the proper frequencies. A standard clamp-on grounding meter, for example, makes measurements at very low frequencies while the electrical frequencies involved in lightning are relatively high. The impedance of the system at the frequencies measured by the device may not be relevant at the frequencies where the performance is needed. NEETRAC, a Center of Georgia Tech, is currently doing studies on these scenarios – including the new intelligent meters being proposed. Although, these results have not been released to the public, representatives from NEETRAC have indicated that it is likely that they will share portions of the results with the IEEE PES Surge Protective Devices Committee which is planning to create a working group paper on these devices via Working Group 3.6.6.

With regard to grounding, bonding, application of SPDs and connection length; I would highly recommend a review of IEEE Standard C62.72(TM)- 2007 and, of course, Clause 4 of IEEE Standard 1100-2005. Clause 11 of C62.72(TM)-2007 speaks to a number of these grounding and bonding issues in detail. Further, Clause 16 provides guidance on installation and coordination issues such as connection lead length.

I apologize for being lengthy in this post; however, this post is brief compared to the time an exhaustive review of this topic would require.

One last note, another interesting paper that deals not only with SPD lead lengths but also loop geometries in the connection of SPDs that one might find useful is "Lingering Lead Length Legacies in Surge- Protective Devices Applications" by François D. Martzloff and Kermit Phipps.

Thanks for reading!


Ron Hotchkiss

Forum Core Group Member, Surge Suppression

Lighting Protection Group: http://tech.groups.yahoo.com/group/LightningProtection

 

Mike Holt’s Comment: Thank you Mr. Hotchkiss for helping us (the electrical professionals) better understand the importance of “bonding” for lightning and surge protection. You confirm what I’ve been trying to say to the electrical industry for many years.

 

Click here to post a comment
[ View More Newsletters ] [ Send to a Friend ] [ Post Comments | View Comments | Notify Me When Comments Are Added ]

Copyright © Mike Holt Enterprises, Inc. All rights reserved.
This article is protected by United States copyright and other intellectual property laws and may not be
displayed or published on the internet without the prior written permission of Mike Holt Enterprises, Inc.

http://www.MikeHolt.com     1-888-NEC-CODE (1-888-632-2633)

Experiencing a Problem? Click Here

 
Comments
  • Field testing the ground rods, with the clamp on meter not being capable of reading higher frequencies, does or would a fall of potential work better or no different, if using a clamp on meter would lower test readings 3 ohms or less help matters at all.

    Pat Pappas
    Reply to this comment

  • This article is not only instructive and accurate, it shows writing skills generally reserved for English teachers. It is a pleasure to read such well presented technical material.

    Thank you for this fine contribution.

    tomstratton
    Reply to this comment

  • Interesting article -- I was surprised when recently I was on-line searching for some information and came across the subject of lighting strikes. I followed the information to some length and ended up with the Patten office reference on lighting strikes.

    I was again surprised at how many pattens have been filed on systems for mitigating lighting strike damage, especially during the early 1900 hundreds. This was a big deal in the early years even before we began to electrify building -- lighting still does cause many fires, we just don't hear about them as often.

    JG

    John Gary
    Reply to this comment

  • The clamp on ground resistance meters use the multiple grounds including the water mains and a multigrounded primary neutral to create the reference ground. This only works if the secondary neutral is bonded to the primary neutral.

    If the primary system is 3-wire 3-phase ungrounded ( e.g. 4,800 volts delta ) or if it is a unigrounded system ( typically 6,640Y11,500 or 6,900Y12,000 ) where the substation grounded neutral is not brought out to the distribution transformers for use as an equipment ground, then the reference ground is only the secondary grounding electrodes at other buildings that are on the same distribution transformer.

    In the case of some industrial plants, the primary neutral stops at the primary metering point. Downstream, all of the transformer primaries are connected phase-to-phase with the earth serving as the primary system equipment ground. One reason for doing this is that some plants need to impress direct current ( cathodic corrosion protection ) onto the grounding electrodes to protect them from a severly corrosive environment. The water service also has a dielectric union for the same reason to prevent diversion of the direct current. For solidly grounded voltages of 6,640Y11,500 volts and up, the earth can serve as an adequate equipment ground particularly for distribution transformers.

    Another instance of where the primary neutral had to be cut was a drive-in theater near Barberton, Ohio where the ground mat at their little 500 KVA industrial substation was actually better the the one at Ohio Edison's transmission station. About once week, the 80,000Y138,000 volt transmission grid would have a flashover - the fault would head right for the drive-in, blow up 3 or 4 speakers, and then head back through the overhead neutral for the 13,200Y23,000 volt industrial distribution.

    I have also read about a plant that used water tanks as a primitive and very rugged variable resistors. What they had to do was to make a cut in the drain pipe and otherwise isolate the water tanks from ground to keep stray current from knocking people out of their bathtubs.

    Michael R. Cole
    Reply to this comment

  • This is an excellent article from someone that has clear knowledge of the physics of lightning. If I extrapolate Ron’s comments regarding the subtle differences between bonding and grounding resistance and impedance, I start to conclude that as we lower the impedance to ground we increase the current demand on the bonding conductors and connections.

    This brings me to the thought that the NEC minimum installation requirements for bonding are relevant for a particular range of impedances and installation characteristics. I wonder how relevant the current NEC minimum standards are to the increasingly lower ground resistances. The NEC is based on years and years of experience. As the base of experience changes, we are only slowly able to identify them and effect a relevant change in the NEC standards.

    The take away I have is that each installation has characteristics that make one or another decision in how to best connect bonds important. The key is to balance the various components so that no component is stressed to the point of failure. Lowering the impedance to ground increases the demands on the bonding conductors in ways that are likely not addressed by the minimum standards in the NEC.

    An argument could be made to control the impedance to ground so it does not fall below a minimum and stess the conductors, connections, or concrete foundations to the point of failure. This is something the NEC has not addressed in the past. Realistically I don’t think it should, but the thought that we would balance our designs in this particular way is something new to me.

    Ken Lillemo
    Reply to this comment

  • I have used the Clamp-on method for years to measure telephone central offices ground fields of all types. However, if the meter is inside a loop, my meter will go to the default setting of 0.7 ohms. The ground loop will have to be corrected to read the true earth resistance of the ground field in question. An example would be a violation of the NEC whereas the neutral is bonded somewhere other than at the Main Service Disconnect. In addition, I have used the three point method to compare the clamp-on method and found a 2 ohm or less difference between the two methods. The clamp-on test set can also be used to measure the isolated line up of a telephone system in central offices. If the tests set reads a high resistance, then you know the line up is isolated, if not, the meter will read a low resistance or go to default, telling you that a ground loop exists somewhere within the line up.

    Curtis Leary
    Reply to this comment

  • This has to be the finest and most concise presentation on many fronts. The posting is rich in significant details that reach far beyond just grounding and bonding issues.

    Sam
    Reply to this comment


Get notified when new comments are posted here
* Your Email:
 
        
 
Add Your Comments to this Newsletter
* Your Name:
   Your name will appear under your comments.

* Your Email:
   Your email address is not displayed.
* Comments:

Email Notification Options:
Notify me when a reply is posted to this comment
Notify me whenever a comment is posted to this newsletter