I&C Engineers vs NEC 250 and 645

[tom baker]

Lets get back to the original question,

645.15 Equipment Grounding and Bonding. All exposed non–current-carrying metal parts of an information technology system shall be bonded to the equipment grounding conductor in accordance with Parts I, V, VI, VII, and VIII of Article 250...

As this is a separate structure, we are all in agreement that a grounding electrode must be added. The question is, should the Isolated ground in the new enclosure be bonded to the new/closest ground rod. or remain isolated for 100's of feet first to a isolated bus in a IT only building then to the isolated triad.

It is my belief that the code requires the initial IG to GEC/EGC bond to be at the new enclosure. If I am wrong, I would like to know and to learn more.

Here is a link to an article in Electrical Contractor https://www.ecmag.com/section/codes-standards/grounding-and-bonding-requirements-nec

Now lets say the hot, for whatever reason, comes in contact with the Isolated Ground. There is no "Low Impendence Path" back to the source. It will travel thru the IG network, to the IG triad. How many ohms between the triad, and the GEC at the transformer feeding the power panel? I am willing to bet it is a lot, MUCH more than 25 ohms. Lets do some simple math I stole from one of Mike Holts books. E/IR 120V 25 ohms = 5A. The OCPD will NEVER trip. So Sad Too Bad Your Dead. Hope he had good life insurance. That is why I believe that a bond is required at the new RIO DCS panel. To trip the feeder breaker.

The purpose statement of the NEC, section 90.1(A) states, “The purpose of this Code is the practical safeguarding of persons and property from hazards arising from the use of electricity.” The NEC DOES NOT CARE about noise on electronic equipment. It is all about life safety.

Yes that seems correct, you must have a low impedance EGC path back to the source. The typical IG used for receptacles with the triangle can terminate at feeder panels, sep derived systems or the service enclosure.

 

[tom baker]

Re: Information Technology
Per the definition, this DCS system, extending to many buildings, is probably not an IT System
645.1 Scope. This article covers equipment, power-supply wiring, equipment interconnecting wiring, and grounding of information technology equipment and systems in an information technology equipment room.
Information Technology Equipment Room. A room within the information technology equipment area that contains the information technology equipment.
If all the wiring in the IT complies with 645 then the IT room is a giant exception to the NEC.
645.4 Special Requirements for Information Technology Equipment Room. The alternative wiring methods to Chap‐ ter 3 and Parts I and III of Article 725 for signaling wiring and Parts I and V of Article 770 for optical fiber cabling shall be permitted where all of the following conditions are met:
(lists 6 conditions that must be met)
I was at a flight trainer for Boeing in Renton, it was an IT Room and had a big red button by the door to shut down the HVAC, also IT rooms usually have raised floors.
If the area does not meet the requirements in 645.4 then its not an IT room and the other articles in the NEC apply

 

[kenaslan]

Re: Information Technology
Per the definition, this DCS system, extending to many buildings, is probably not an IT System
645.1 Scope. This article covers equipment, power-supply wiring, equipment interconnecting wiring, and grounding of information technology equipment and systems in an information technology equipment room.
Information Technology Equipment Room. A room within the information technology equipment area that contains the information technology equipment.
If all the wiring in the IT complies with 645.4 Special Requirements for Information Technology Equipment Room. The alternative wiring methods to Chap‐ ter 3 and Parts I and III of Article 725 for signaling wiring and Parts I and V of Article 770 for optical fiber cabling shall be permitted where all of the following conditions are met:

Agreed. The DCS is outside so I agree, it does not come under the jurisdiction of 645. The room does. (The room is a dedicated IT building, and is not part of this scope. It is existing) But that is not my question. How do I convince the I&C Engineer, that the Isolated Ground "4/0 Cu Drain Wire" Must be bonded in the field mounted Hoffman Box that encloses the DCS? The code is quite clear IMO that ALL non-current carrying metal must be bonded. Unless I am misreading the Code.

 

[petersonra]

He probably does not need to bond it with the 4/0 wire.

Whatever EGC comes into the box is adequate for NEC purposes. For instance, if they used metal conduit, that would be adequate.

 

[don_resqcapt19]

645.15 Equipment Grounding and Bonding. All exposed non–current-carrying metal parts of an information technology system shall be bonded to the equipment grounding conductor in accordance with Parts I, V, VI, VII, and VIII of Article 250...

In my case, the Isolated ground goes 800 feet to a IG bus, then to a triad another 100 feet away. That tiad is connected to the buried ground next to it. The panel supplying the 120V to the DCS is 400 feet in the other direction. We now have a ground loop. The Isolated Ground is to "drain" noise away from the electronics. However my limited understanding is that the loop will cause more problems. In addition there will be a difference in impedance between the two grounds and a nearby lighting strike will cause problems due to the difference in grounds. I believe that the IG should bond to the EGC at the panel, AND also be bonded to a ground rod driven at the location that is also tied to the grounding pad. This would be just like the requirements that a ground rod be driven when you bring power to a separate building, even though you have a EGC going back to the panel. NEC 250.32(A)
I ask as this is such a muddled part of the code.

Typically on DCS equipment, the isolated ground is not used as an EGC and is not part of the fault clearing process. It is used for the shields of shielded cable and for the grounding of the low voltage side of the equipment. There is no ground loop because the EGC and the isolated ground conductor are not connected together at any point except where you say the isolated ground triad is connected to the buried ground next to it. That bonding of the two grounding systems should eliminate the difference in impedance as the bonding results in a single grounding electrode system.
I don't see a code issue and while your distances are longer than normal, the installation is exactly what was required by Emerson for the last DeltaV DCS system I installed.

Note that if you are using the isolated grounding conductor as the fault clearing path, the answer would be totally different as in that case the isolated grounding conductor would originate at the location of the system bonding jumper if the source is an separately derived system or at the main bonding jumper if the source is a service.

 

[Wendtcw]

Does anyone have a good argument besides "that's what the Code states and that is how we will do it" To convince an I&C engineer that the isolated ground in a DCS/PLC/Instrument cabinet MUST be bonded to the EGC. I even had a PE tell me that as long as you go to a ground rod if there is a short, it will trip the breaker. After a quick refresher on ohms law did he see the error. But alas, how do I convince an Instrument engineer to properly bond their equipment. Even the drawings from a manufacture insist on no neutral to ground bond XO is neutral, and the case goes to a ground rod. Never to be connected. They don't seem to care what the NEC says.

I am an Sr Automation (aka, I&C) Engineer and a PE in several states. A long time ago, before 1980's, I found that isolated grounding (using an independent ground rod) of DCS systems was common practice, right or wrong. It was thought that this would minimize electrical noise. It has been a long time since I've seen this in current DCS/PLC implementations. Isolated ground (IG) receptacles (using a separate ground wire back to the electrical service), were popular for a short time, until it was realized that they could cause voltage differences between equipment and is no longer recommended for computer/DCS/PLC systems.

Obviously, saying the NEC requires (not suggests) it should settle the issue. Pointing out the potential for ground loops is reduced with more (mesh) grounding might be a talking point. Also, the best practice standard, TIA-607 Generic Telecommunications Bonding and Grounding for Customer Premises, describes the Bonding of of Rack and computer rooms which in many respects is more stringent (i.e. bigger bonding conductors) than the NEC requires. Within TIA-607, the electrical and telecommunication systems are shown bonded together, as you would expect.

 

[augie47]

Think a picture will help ?