Mike,

I hope all things are well with you. This photo was taken of a control cabinet we are to hook up on one of our projects. Note that the instructions on the top left side indicate that the equipment must be grounded to a ground rod or a ground wire from the supply service.

Naturally, a ground rod is not suitable to serve as the low-impedance fault-current path to the source necessary to clear a ground fault. I thought you might find this humorous.

G. Scott Harding
President, F. B. Harding, Inc.
301-315-0900
gsh@fbharding.com

Mike Holt's Comment: Scott (electrical contractor/engineer) is serving his first term on Code-Making Panel No. 5 - Article 250 Grounding and Bonding. I am very impressed with his desire to help the electrical industry and I look forward to his contributions. Plus, I think he's really a great guy and personal friend.

Currently the NEC in Article 100 defines the terms "ground" or "grounded" as "connected to the earth or to some conducting body that serves in place of the earth." Yet, the NEC often uses the term "ground" when it really means "bond" (connected to an effective ground-fault path to clear a fault) [250.2 and 250.4(A)(5)].

Grounding - Grounding metal parts to the earth in premises wiring is only useful to provide a path for lightning, shunting high-frequency noise, or reducing static discharge.

Bonding - Bonding all metal parts together and then to the system winding (typically to the X0 terminal of a transformer) is done to provide a low-impedance path to the source (system) to facilitate the opening of the circuit-protection device to remove dangerous voltage on metal parts. In addition, bonding the system to metal parts (typically to the X0 terminal of a transformer) stabilizes the system voltage to the metal parts and it provides a zero system reference (to the metal parts).

It took me over 20 years to figure out that when the NEC, as well as books and magazine articles, state "ground or grounded" they rarely intend that the metal parts be connected to the earth or to some conducting body that serves in place of the earth. Typically, the intent is to have the metal parts "bonded" to a low-impedance path so that fault current can return to the system winding. See the definition of "effective ground-fault path" in 250.2 and review 250.4(A)(3), (4), and (5).

For example: Section 250.86 in the NEC requires metal enclosures and raceways to be "grounded."

Is it really the intent of this Code section that metal enclosures and raceways be grounded to the earth (like with a ground rod)? No, the metal raceway ultimately needs to be bonded to the system winding (transformer or generator), by bonding the metal parts to the "effective ground-fault current path," as defined in 250.2.

I'm very concerned when the term "ground," "grounded, or "grounding" is used to refer to the practice of clearing a line-to-case fault (bonding), because some actually think that connecting metal parts to the earth will serve to assist in clearing a ground-fault, which is false for systems that operate at 600V or less [250.4(A)(5)].

I hope in the future, when the NEC, as well as writers of books and magazine articles, refer to making an electrical connection to ensure that the protective device opens to clear a ground fault, that it's clear. Any reference to "ground" or "grounded" in a book or magazine should only be used when the intent is to make a connection to the earth or to some conducting body in place of earth (like a ground rod).

Ground Summary
For premises' systems that operate under 600V, grounding "does not":

• Assist in clearing a line-to-case fault because of it's high resistance (IEEE 142 - 2.1.4 and 4.1.1)
• Establish a "zero system reference"
• Reduce shock hazard to a safe value (IEEE 142 - 2.2.8)

Grounding does provide a path to the earth for:

• Lightning
• Static electricity
• High-frequency noise

Bonding Summary
For premises' systems that operate under 600V, bonding the system to metal parts does:

• Stabilize the system voltage to referenced metal parts [IEEE 142 - 1.5.2.]
• Assist in clearing a line-to-case fault [250.4(A)(3), (4) and (5)]
• Establish a "zero system reference" (IEEE 142 - 2.2.8)

We as an industry need to collectively come together to clarify the purpose and usefulness of grounding versus bonding.

NEC Task Group.
A NEC task group is to be formed to evaluate the use of terms in the NEC. However, the recommendations from this task group are for the 2008 NEC, so don't expect to see any significant changes on this issue for the 2005 NEC. Considerations should be given to coordinate the use of ground and bond with the Canadian Code. For example:

Our "Part VI. Equipment Grounding and Equipment Grounding Conductors," in the Canadian Code is "Equipment Bonding."

Our "250.119 Types of Equipment Grounding Conductors," in the Canadian Code is "10-804 Material for Bonding Conductor."

Our 250.86, in part "IV. Enclosure, Raceway, and Service Cable Grounding," states: "Except as permitted by 250.112(I), metal enclosures and raceways for other than service conductors shall be grounded." Yet in the Canadian Code, "Conductor Enclosure Bonding" states: "Metal enclosure for conductors shall be bonded to ground" [10-302].

Now I don't like the Canadian Code's use of "bonding to ground," but the Canadian Code does use bonding when the intent is bonding. At least they are much further ahead of us.

Why all the confusion?
I have done a lot of thinking on this question and I think it relates to the historical practice of using terms and concepts developed a hundred years ago from the electric utilities. Actually, if you use the terms [Article 100] and concepts [250.4(A)(1)] contained in the NEC, they are 100 percent correct as they relate to multipoint-grounded high-voltage systems, but not always for under 600V systems.

Many manufacturers of grounding fitting stress that the earth is the greatest thing for electrical safety and the more of it you used, the better. Plus, the copper guys really would like you to use lots of larger copper wire. Some manufacturers claim that high-voltage/high-current flow from lightning will cause concrete to crack, so you'd better ground the steel in concrete to a ground rod via a 6 AWG copper wire to solve this problem. This is a false statement.

It's sad, but electrical engineers are rarely exposed to the topic of grounding or bonding as it relates to systems that operate under 600V in college. So, when they come out into the real world, they really don't understand these systems. So who's there to save the day? The salespersons of grounding fitting manufacturers lobby the electrical engineers' community that the earth is so wonderful and a "good ground" is essential to prevent almost everything imaginable (snake oil, if you ask me).

Sadly over-engineering of the grounding system, which doesn't provide any value to the customer, is costing property owners millions of dollars. One clue that someone doesn't really understand grounding versus bonding, is when a job runs over budget, they decide to "cut" some of the grounding design. How is it possible to "cut" something that was necessary? It's either required/necessary, or it's not.

Now, I'm not saying that all engineers, salesmen and manufacturers of grounding fitting don't understand the differences between grounding and bonding. I bet there are lots of individuals reading this newsletter that thought they must be crazy, because they couldn't understand the value of the grounding system design or manufacturer claims.

How can you help?
If you are concerned about the misuse of the term "ground," "grounded," or "grounding," please - when you see a magazine article or book misuse these terms, send a letter to the editor asking them to clarify the intent of the article. Each of us has the responsibility and opportunity to clarify this issue, and I think if we work together, we will make this happen.

Can you help me?
For those who design electrical systems, can you answer me the following questions.
Ground Ring.

• For other than lightning, what is the purpose of a ground ring?
• For other than lightning, when would a ground ring be installed?
• How do you determine the conductor size for the ground ring?
• Are ground rods attached to the ring and why?
• Do you and how often do you bond the ring to the structural steel of a building?
• What standard did you use to answer the above ground ring questions?

Ground Rod. Why purpose would a ground rod service if it were installed at any of following locations?

• Machine tool (CNC machine)
• Steel column in a building
• Metal pole for luminaires
• Motor
• Street lighting
• Metal pole or metal parts for traffic signal

If you have any technical documents or standards, like an IEEE paper, that recommends any of the above actions, please E-mail it to Mike@MikeHolt.com or fax it to 954.720.7944.

Please don't support your position with what you think, what you were told, or what was written in a magazine or book. Magazine articles and books are often based on what the author was told, what he/she read, or what he/she thought, sometimes without researching the facts. How do I know? Looking back now, I see I committed these acts, and for that I apologize for my contribution to the confusion on grounding versus bonding.

God Bless,

P.S. A wise person once told me that you'll become highly educated by getting the answers to your questions, but you'll make discoveries (maybe not a lot of friends) by questioning the answers you receive. Just wanted to leave you with that thought.